Territorial Cohesion Discourse: Key Concepts from a Geographical Perspective

MASARYK UNIVERSITY FACULTY OF SCIENCE DEPARTMENT OF GEOGRAPHY

Territorial cohesion discourse: key concepts from a geographical perspective

Ph.D. Dissertation

Jiří Malý

Supervisor: Mgr. Ondřej Mulíček, Ph.D. Brno 2016

Bibliographic Entry

Author: Mgr. Jiří Malý Faculty of Science, Masaryk University Department of Geography

Title of Dissertation: Territorial cohesion discourse: key concepts from a geographical perspective

Degree Programme: Geography

Field of Study: Regional Geography and Regional Development

Supervisor: Mgr. Ondřej Mulíček, Ph.D. Faculty of Science, Masaryk University Department of Geography

Academic Year: 2016/2017

Number of Pages: 194

Keywords: territorial cohesion; spatial disparities; urban systems; polycentric development; accessibility; services of general interest

Bibliografický záznam

Autor: Mgr. Jiří Malý Přírodovědecká fakulta, Masarykova univerzita Geografický ústav

Název práce: Diskurz územní soudržnosti: klíčové koncepty v geografické perspektivě

Studijní program: Geografie

Studijní obor: Regionální geografie a regionální rozvoj

Školitel: Mgr. Ondřej Mulíček, Ph.D. Přírodovědecká fakulta, Masarykova univerzita Geografický ústav

Akademický rok: 2016/2017

Počet stran: 194

Klíčová slova v češtině: územní soudržnost; prostorové nerovnosti; sídelní systémy; polycentrický rozvoj; dostupnost; služby obecného zájmu

Abstract This dissertation focuses on the topic of territorial cohesion. Although territorial cohesion is one of the objectives of EU cohesion policy, its meaning is still ambiguous and vague. Inconsistent understanding and interpretation of the concept, leading to the multi-interpretative character of territorial cohesion, is reflected in the current discussions about its true essence. Besides the political level, the territorial cohesion discourse is tightly connected with the non- binding EU spatial planning platform. In this context, the objectives of polycentric development and fair access to basic services, among others, are proclaimed in relation to territorial cohesion. However, the efforts to establish universal planning tools are facing several problems stemming from diverse spatial configurations of national/regional urban systems, different traditions of spatial planning and, not least, varied geographical scale levels at which principles of territorial cohesion are applied. The aim of this study is thus to identify the key principles and concepts of the territorial cohesion discourse and then evaluate the spatial reality of current urban systems functioning, which is interpreted in the context of the EU spatial planning narratives. The assumption is a discrepancy between the idealized imagination of balanced spatial development and the actual socio-spatial processes, primarily due to the increasingly complex structure of relational space affecting the shape of urban networks. Urban systems of the Czech Republic and the South Moravian Region are analysed using quantitative methods of geographical research. Emphasis is placed primarily on the analysis of polycentric development and its relation to the level of territorial disparities, and the issue of access to services of general interest with regard to the role of small towns within a broader urban system. The results of the empirical part show that the principles of territorial cohesion, as a representation of space, are not in line with contemporary trends and dynamics of urban systems, and their application in a specific territory must respect its unique conditions. The study concludes that territorial cohesion, as a principally political concept, is very difficult to operationalize at the level of spatial planning, with the main reason being the absence of empirical evidence proving the beneficial effect of its fundamental principles.

Abstrakt Disertační práce se zabývá tématem územní soudržnosti. Ačkoliv je územní soudržnost jedním z cílů Kohezní politiky EU, její význam je stále nejednoznačný a vágní. Nejednotné chápání a interpretace konceptu, vytvářející diskurzivní povahu územní soudržnosti, se promítají do probíhajících diskuzí o jeho skutečném významu. Kromě politické roviny je normativ územní soudržnosti nedílnou součástí nezávazné platformy prostorového plánování na úrovni EU. V tomto kontextu jsou v souvislosti s územní soudržností proklamovány, mimo jiné, cíle polycentrického rozvoje a spravedlivého přístupu k základním službám. Snahy o vytváření univerzálních plánovacích nástrojů však naráží na řadu problémů pramenících z odlišných prostorových konfigurací národních/regionálních sídelních systémů, odlišných tradic územního/prostorového plánování a v neposlední řadě odlišných měřítkových úrovní, ve kterých jsou principy územní soudržnosti uplatňovány. Cílem této práce je tak nejprve identifikace klíčových principů územní soudržnosti jako normativního konceptu a poté zhodnocení prostorové reality fungování současných městských systémů a jejich interpretace ve světle narativu územní soudržnosti. Předpokladem je nesoulad mezi idealizovanou představou územně vyváženého rozvoje a skutečnými sociálně prostorovými procesy, a to především z důvodu stále komplikovanější struktury relačního prostoru ovlivňující podobu městských sítí. Pomocí kvantitativních metod geografického výzkumu je analyzován sídelní systém České republiky a Jihomoravského kraje. Důraz je kladen zejména na analýzu polycentrického rozvoje a jeho vztah s mírou územních disparit a problematiku dostupnosti služeb obecného zájmu v souvislosti s rolí malých měst v systému osídlení. Výsledky empirické části ukazují, že principy územní soudržnosti, jako reprezentace prostoru, se poměrně značně rozchází s aktuálními trendy dynamiky sídelních systémů a jejich aplikace na konkrétní území musí respektovat jeho specifické podmínky. Studie se přiklání k závěru, že územní soudržnost, jako primárně politický koncept, je velmi obtížné operacionalizovat v rovině prostorově plánovací, přičemž hlavní příčinou je absence empirických důkazů o všespásné roli jejích základních principů.

Prohlášení

Prohlašuji, že v souladu s § 47b zákona č. 111/1998 Sb. v platném znění souhlasím se zveřejněním své disertační práce fakultou, a to v nezkrácené podobě elektronickou cestou ve veřejně přístupné části databáze Informačního systému Masarykovy univerzity. Prohlašuji, že jsem disertační práci vypracoval samostatně s použitím pramenů a literatury uvedených v seznamu citované literatury.

V Brně, dne 30. listopadu 2016 ……………………………….. Jiří Malý

Acknowledgement At this point I would like to thank all those who have significantly contributed to writing the presented dissertation and without whom the thesis might not have been completed. Many thanks go to my supervisor Mgr. Ondřej Mulíček, Ph.D. whose professional guidance and advices, provided during my entire study, have become an invaluable source of inspiration for my research. At the same time, I thank him for human approach and permanent willingness to discuss not only the geographical issues. I would also like to thank a colleague from the Institute of Geonics of the CAS RNDr. Bohumil Frantál, Ph.D. for helping me with preparation of individual papers and for providing expert recommendations in the field of statistical analyses. In addition, I drew inspiration from discussions with experienced geographers from the Department of Geography at MU, especially with RNDr. Robert Osman, Ph.D., Mgr. Jaroslav Biolek and Mgr. Daniel Seidenglanz, Ph.D. The closest family has been very important element during the whole research and work on the dissertation. Above all, I thank my wife Linda for initial impulse, motivation, patience and tolerance. Special thanks go to my mother and my late father for their endless support throughout my long-standing study period. I also thank the Hanousek family for pleasant living conditions which they provided to me at the beginning of my studies and my brother for introducing me to the world of geography.

Poděkování Na tomto místě bych velice rád poděkoval všem, kteří výrazným způsobem napomohli sepsání předkládané disertační práce a bez nichž by tato práce možná ani nebyla dokončena. Velký dík zaslouží můj školitel Mgr. Ondřej Mulíček, Ph.D., jehož odborné vedení a rady, poskytované nepřetržitě během celého studia, se staly neocenitelným zdrojem inspirace pro mé bádání. Zároveň mu děkuji za lidský přístup a konstantní ochotu diskutovat nejen nad geografickými tématy. Dále bych chtěl poděkovat kolegovi z Ústavu Geoniky AV ČR RNDr. Bohumilovi Frantálovi, Ph.D. za odbornou pomoc při vytváření jednotlivých článků a za cenná doporučení v oblasti statistických analýz. Inspirující pak pro mě byly rovněž odborné debaty se zkušenými geografy z Geografického ústavu MU, zejména RNDr. Robertem Osmanem, Ph.D., Mgr. Jaroslavem Biolkem a Mgr. Danielem Seidenglanzem, Ph.D. Po celou dobu výzkumu a sepisování disertační práce mi byla důležitou oporou nejbližší rodina. Především děkuji manželce Lindě za prvotní impuls, motivaci, trpělivost a toleranci. Obrovský dík patří mamince a zesnulému tatínkovi za jejich neutuchající podporu v průběhu mých dlouhých studijních let. Děkuji také rodině Hanousků za poskytnutí příjemných životních podmínek v počátcích mého studia a bratrovi za to, že mě seznámil se světem geografie.

CONTENTS 1 INTRODUCTION ...... 19 1.1 Aims, objectives and research questions ...... 21 1.2 Structure of the dissertation ...... 22 2 TERRITORIAL COHESION: POLICY AND PLANNING CONTEXT ...... 25 2.1 EU cohesion policy ...... 25 2.2 From common market to territorial cohesion ...... 26 2.3 Territorial cohesion as a part of EU cohesion policy ...... 29 2.4 Specification of the concept of territorial cohesion ...... 31 2.4.1 The role of ‘European spatial planning’ ...... 31 2.4.2 Multidimensionality of territorial cohesion ...... 33 2.5 Territorial cohesion as a discourse ...... 38 2.5.1 The roots of territorial cohesion ...... 39 2.5.2 Territorial cohesion ‘storylines’ ...... 41 2.5.3 Issues of cohesion and competitiveness ...... 44 2.5.4 Polycentric development ...... 46 3 NARROWING DOWN TERRITORIAL COHESION COMPLEXITY ...... 49 3.1 Territorial cohesion as a path to balanced development ...... 49 3.2 Polycentricity and access to SGI ...... 50 3.3 The issue of scale ...... 50 4 STUDY AREA AND METHODOLOGICAL OVERVIEW ...... 53 5 EMPIRICAL ANALYSIS OF TERRITORIAL COHESION ...... 57 Paper 1: European territorial cohesion policies: Parallels to socialist central planning? ..... 57 Paper 2: Moving towards more cohesive and polycentric spatial patterns? Evidence from the Czech Republic ...... 77 Paper 3: Impact of polycentric urban systems on intra-regional disparities: A micro-regional approach ...... 95 Paper 4: Small towns in the context of ‘borrowed size’ and ‘agglomeration shadow’ debates: The case of the South Moravian Region (Czech Republic) ...... 117 Paper 5: Questioning territorial cohesion: (Un)equal access to services of general interest ...... 133 6 CONCLUSIONS ...... 155 7 ZÁVĚRY ...... 159 REFERENCES ...... 163

FIGURES ...... 183 TABLES ...... 185 LIST OF ABBREVIATIONS ...... 187 APPENDICES ...... 189

1 INTRODUCTION

‘Perfectly even development, complete socio-spatial equality, pure distributional justice, as well as universal human rights are never achievable. Every geography in which we live has some degree of injustice embedded in it, making the selection of sites of intervention a crucial decision.’ (Soja, 2009, p. 3)

Achieving territorial cohesion is one of the fundamental objectives of the European Union (EU) regional policy. Until recently, the EU regional policy has primarily focused on overcoming economic and social disparities among EU society, thus aiming to strengthen economic and social cohesion. In recent years, however, the economic and social cohesion has been enriched by a third dimension: territorial (EC, 2007b). Consequently, the EU regional policy is also referred to as cohesion policy including aspects of economic, social and territorial cohesion. The basic objective of economic cohesion is sustainable economic growth and economic convergence ensuring the global competitiveness of EU territory. Social cohesion represents an effort to reduce social disparities, eliminate social exclusion and strengthen social ties across the current society (Farrugia and Galina, 2008). From a very simplified and generalized perspective, territorial cohesion can be considered a territorial dimension of the objectives of economic and social cohesion emphasizing balanced spatial development and combating territorial disparities. In fact, however, the definability of all three dimensions of cohesion policy tends to be very difficult. Moreover, the conceptualization of the territorial dimension is still the most ambiguous. The EU regional policy has been a part of EU policies for decades and from the beginning, its main interest has been to assist less developed regions within the European Community (EC)/EU. Although EU is typically ranked among economically and socially most developed areas of the world, the internal differences between individual EU regions are huge. The enlargement of the EU with 13 new member states between 2004 and 2013 have increased regional disparities and strengthened the role of cohesion policy. The concentration of economic activities primarily into urban areas and areas strategically located in close proximity to transport infrastructure exceeds the concentration of population. Such areas are characterised by good access to basic services such as healthcare, education facilities or business services, and offer a sufficient number of jobs (however, economic growth is in many cases accompanied by issues of environmental burden, congestion or social exclusion). On the other hand, there are regions that are far from reaching such high levels of service accessibility and are distinguished by specific structural handicaps with related social problems. From the EU point of view, the unfavourable situation of disadvantaged regions should be solved by territorial cohesion policies aiming at the elimination of spatial inequalities. Based on the EU documents focusing on territorial cohesion (e.g. EC, 2008; ESPON, 2012a), it can be suggested that territorial cohesion should be achieved at all geographical scales, i.e. in all EU territories. However, the scale itself affects the meaning and the form of territorial cohesion. Thus, territorial cohesion is strongly linked with and influenced by

19 geographical scale. While networking of cities in terms of business cooperation and transport connections contributes to increasing economic performance of the largest agglomerations and consequently improves cohesion at the EU level, territorial cohesion of a particular locality (town district, neighbourhood) is more tightly related to the issues of spatial justice (see Soja, 2010) and would be based on completely different factors including e.g. social aspects (social segregation, exclusion, heterosexual bias, patriarchy), demographic situation (age structure), health conditions (disability) or the issues of governance and local development. Territorial cohesion should be therefore treated as a scale-dependent concept. Besides the issue of geographical scale, difficult definability of the concept stems from its strong political dimension. Advocates of diverse political convictions perceive territorial cohesion differently. Consequently, a number of interpretations contribute to the fuzziness and excessive complexity of the concept (see e.g. Zonneveld and Waterhout, 2005; Camagni, 2007; Davoudi, 2007). However, all of them subscribe to the notion of territorial cohesion and therefore contribute to the institutional embeddedness of territorial cohesion discourse. In fact, the position of territorial cohesion within the prevailing neo-liberal economic model of the EU territory gives rise to doubts and speculations about the meaning of territorial cohesion. In that instance, territorial cohesion is being interpreted as a concept characterized by contradictory objectives and incomprehensible meaning (see e.g. Faludi, 2005a; Doucet, 2006; Servillo, 2010). One of the most discussed issues is the tension between the dominant model of economic development accentuating competitiveness and the goal of cohesion based rather on social solidarity principles (Davoudi, 2005; Hudson, 2015). Political normative narratives underlying the territorial cohesion discourse are transformed into more specific spatial planning goals and concepts including namely polycentric development and equal access to services of general interest with emphasis placed on the role local towns play in the development of rural and peripheral areas (EU Ministers responsible for Spatial Development, 2011). The inclusion of territorial cohesion principles into the ‘non-binding’ EU spatial planning platform results in even stronger integration of the concept in the EU regional policy. Nevertheless, recent studies have shown that the support of territorial cohesion and its key goals (polycentric development in particular) is not based on sufficient amount of empirical research (Meijers and Sandberg, 2006; Veneri and Burgalassi, 2012) and should be the subject of further examination. The dissertation seeks to draw attention to some of the aspects of confusion that enshrouds territorial cohesion; it also attempts to deeply understand the key normative concepts of territorial cohesion through geographic and more analytical perspective. The scope and primarily the purpose of the dissertation do not allow to examine territorial cohesion in all its complexity (which is even not practically possible for such complex umbrella concepts), and hence the research is focused only on relevant aspects characterized by crucial importance from the perspective of the regional dimension of territorial cohesion. The following sections present the main aim and the objectives of the presented study and outline the structure of the thesis.

1.1 Aims, objectives and research questions The topic of the dissertation is the idea of territorial cohesion representing the territorial dimension of EU cohesion policy. The central focus is put on a regional dimension of territorial cohesion in the context of socioeconomic spatial inequalities acquiring complex and various forms depending on the structure of urban networks and proximity to socially valued resources and technical infrastructure. The main aim of the dissertation is to deeply understand the crucial normative principles and concepts of territorial cohesion and to evaluate their relation to contemporary spatial development which reflects functional dynamic of urban systems and accessibility of basic services and transport infrastructure. Five objectives are defined within the framework of the main aim. Each of them is based on a predetermined hypothesis and specified by two research questions. The following order of the particular objectives is in accordance with the sequence in which they are analysed: 1) Mutual comparison of territorial cohesion conceptualisations in two different historical periods conducted in order to critically evaluate the novelty of EU territorial cohesion discourse. Hypothesis: Despite the term territorial cohesion has appeared in connection with the contemporary EU regional policy, it resembles in some aspects the central planning doctrines endorsed by socialist regimes (especially in the former Czechoslovakia). Research questions: a) What are the differences in the political discourses framing the concept of territorial cohesion in the two periods (the current EU regional policy and the former socialist regime)? b) What are the main features of the particular spatial policies shaping the functional geographies of interdependent places and territories? 2) Evaluation of recent changes in spatial routines and practices of contemporary society shaping the hierarchy of relational space in the context of the territorial cohesion discourse. Hypothesis: There are discrepancies between specific representations of space stemming from the normative cohesion-policy agenda and the spatial practices imprinted into the development and functioning of the real urban system. Research questions: a) To what extent spontaneously produced settings fit with the normative imagination of polycentric and balanced space? b) Do the ongoing changes make for a more polycentric organization of space? 3) Evaluation of the impact of polycentric urban systems as the key spatial planning goal of the territorial cohesion discourse on equal spatial development. Hypothesis: As the core concept of the territorial cohesion discourse, polycentric development is viewed as a process that contributes to lowering regional disparities and to the overall economic growth of the European territory.

Research questions: a) Is it possible to prove polycentricity has a positive impact on overcoming intra- regional disparities? b) Is the support of polycentricity based on empirical evidence? 4) Deepening the knowledge about the distribution of socially valued resources (labelled as services of general interest within the terminology of the EU regional policy) across the network of small towns in the context of ‘borrowed size’ and ‘agglomeration shadow’ discussions. Hypothesis: The service function of small towns is affected more by the position of a particular town in relation to the core city of the metropolitan region and transport infrastructure than by the population size of the town. Research questions: a) Which factors are associated with the level of the service function of small towns? b) Which towns borrow size and which lie in the agglomeration shadow? 5) Clarification of the impact of spatial differences in access to services of general interest on living conditions of particular areas. Hypothesis: Unequal access to services of general interest plays a significant role in the formation of socio-spatial inequalities. Research questions: a) What is the relationship between the level of access to services of general interest and demographic and socioeconomic conditions? b) How can this relationship be interpreted regarding the structure of an urban system and routing of transport infrastructure? In spite of the formal separation of the particular objectives, the territorial cohesion discourse interweaves all of the research questions. In order to answer these questions, the study is structured into several consecutive parts.

1.2 Structure of the dissertation The dissertation consists of two main parts. The first, theoretical part, is logically divided into particular thematic chapters; the second, empirical part, is composed of five consecutive studies that have been published or accepted for publication (the exception is Paper 2 - under review at the time of writing this dissertation) in international scientific journals in the form of research papers. The theoretical background might be understood as a necessary commentary introducing the whole theme of territorial cohesion and particular issues relevant for the research. The author of the dissertation is also the main author of all the published studies. The character of the presented thesis can be described as both theoretical and empirical. While the theoretical part is predominantly based on a literature review focusing on identification of crucial notions, statements, categories and concepts, the empirical part relies on quantitative approaches operating with statistical data. The ‘hypothetico-deductive’ model of research is characterized by a predetermined sequence of individual tasks which starts with the formulation

22 of basic relations (based on theoretical research), hypotheses and questions, followed by collecting data, conducting analyses, testing hypotheses and making final statements and conclusions. Based on this logic, formulation of research questions actually follows theoretical framework. Nevertheless, as regards formal structure of the presented study, the main aim and research questions are listed in the introduction in order to make the reader more familiar with the purpose of the study in the beginning. The main reason is to prevent the reader from reading a text without being aware of its primary intent. The initial theoretical framework (Chapter 2) stems from the normative character of territorial cohesion as an objective of EU cohesion policy that is based on political proclamations formulated by EU power elites. Territorial cohesion is therefore approached through the normative narrative of EU cohesion policy (Molle, 2007) including the aspects of social solidarity and cohesion (Davoudi, 2007; Peyrony, 2007) and also the support of economic competitiveness (Doucet, 2006; Evers, 2008; Vanolo, 2010). From a policy perspective, the main information sources have become the EU Treaties – the Treaty of Amsterdam (EC, 1997) and the Lisbon Treaty (EC, 2007b); regular Reports on Economic, Social and Territorial Cohesion (EC, 2004a; EC, 2010b); Green Paper on Territorial Cohesion (EC, 2008); and academic publications dealing with cohesion policy (e.g. George, 1996; Blair, 2005; Molle, 2007; Faludi, 2010; Fiala and Pitrová, 2010). Based on the literature review, the key aspects of territorial cohesion are summarized. Firstly (Sections 2.1 and 2.2), the nature of cohesion policy is briefly characterized and its historical development is outlined. Next (Section 2.3), attention is focused on the concept of territorial cohesion, its definability, and meaning in the EU political context. Besides the political perspective (where territorial cohesion originates from), the concept of territorial cohesion is reflected within the EU spatial planning platform which has, however, a non- binding character and only serves as a basis for the creation of EU regional policies (Abrahams, 2013). To understand the importance of territorial cohesion in planning practice, the documents focusing on the priorities of ‘spatial planning’ at EU level – European Spatial Development Perspective (EC, 1999), Territorial Agenda (EU Ministers responsible for Spatial Development, 2007), Territorial Agenda 2020 (EU Ministers responsible for Spatial Development, 2011); studies of the ESPON programme (e.g. ESPON, 2006a, 2012a, 2012c); and expert reports (e.g. Sapir et al., 2004; Evers et al., 2009) have been examined. The aim of Section 2.4 is to interpret the meanings territorial cohesion gains within the framework of EU spatial planning analytical practises. Based on the literature review concerning the political and spatial planning dimensions of territorial cohesion, the discourse nature of the concept is described in Section 2.5. Emphasis is placed on a fundamental dichotomy of cohesion and competitiveness accompanying the territorial cohesion discourse. The study identifies the key objectives of spatial planning including polycentric development (Davoudi, 2003; Hague and Kirk, 2003) and the issue of equal accessibility of services of general interest (Davoudi, 2005; Rauhut and Ludlow, 2013). However, their deeper theoretical and empirical analysis is the subject of the empirical part of the dissertation. Subsequently, the main aspects of the idea of polycentric development that are closely related to the cohesion/competitiveness issue are characterized. The purpose of Chapter 3 is to highlight the key aspects of the current understanding of territorial cohesion and interlink it with the research questions and the main research objective.

The research methods are described in the following Chapter 4. The purpose and nature of regional analyses require the use of quantitative research methods. These are mostly statistical analyses of secondary data. Depending on the research aim (based on the partial research questions), appropriate methods are selected. The empirical part framed as Chapter 5 consists of five studies in the form of research papers; each study focuses on one of the five research objectives. However, individual studies cannot be understood as isolated parts but as an interconnected and coherent work which fulfils the main research aim through the means of specific studies. The contribution of individual authors to the presented papers is described in Appendix 4.

2 TERRITORIAL COHESION: POLICY AND PLANNING CONTEXT

2.1 EU cohesion policy The conceptualization of territorial cohesion is not possible without understanding the causes of its origin and the sense of the cohesion policy, which can be considered the essential framework producing the territorial cohesion discourse. Although the cohesion policy experienced relatively complicated and long evolution, its primary goal remains unchanged. The cohesion policy aims at reducing inequalities between rich and poor regions and mitigating divergent trajectories of socioeconomic development. The support of economic, social and territorial cohesion is one of the major EU goals, which is included in the Lisbon Treaty (also known as the Reform Treaty) from 2007, ratified by all member states in 2009 (EC, 2007b). It contains a provision that cohesion is one of the shared competences between the EU and the member states1. In the Consolidated version of the Treaty, Article 174 states (EC, 2012, p. 127): ‘In order to promote its overall harmonious development, the Union shall develop and pursue its actions leading to the strengthening of its economic, social and territorial cohesion. In particular, the Union shall aim at reducing disparities between the levels of development of the various regions and the backwardness of the least favoured regions.’ Cohesion principles primarily express solidarity between member states and particular EU regions (Skokan, 2008). According to Ahner (2009), cohesion policy is the clearest and the most obvious expression of solidarity between EU member states. Solidarity, however, does not consist only of the compensation payments to disadvantaged regions, but also in a true effort to support their development. What level of solidarity, then, would be sufficient for achieving cohesion? Molle (2007) argues that cohesion is generally understood as a certain level within which social and economic disparities between EU states or regions are socially and politically accepted. Direct assessment of cohesion then should not be based on a static examination of the reached socioeconomic level of the particular territory, but on the dynamic of its development, i.e. the change in disparities during a specified time period. Decline in disparities indicates the strengthening of cohesion (convergence of regions) and, vice versa, growing disparities weaken the cohesion (divergence of regions). EU funds represent the main instrument for implementing the cohesion policy. Their essence lies in the redistribution of funds to the less developed areas of the EU, aiming to reduce economic and social disparities between regions. This should result in economic growth and increased living standards in underdeveloped regions. The principle of solidarity, when economically more developed and richer countries help the less developed ones, is firmly grounded in EU cohesion policy. An integral part of the cohesion policy should be an exchange of experiences and cooperation between regions, consultancy within the EU educational programs, open discussions with all stakeholders and dialogue among all levels of government

1 In the Consolidated version of the Treaty on European Union, Article 2, second paragraph, states: ‘When the Treaties confer on the Union a competence shared with the Member States in a specific area, the Union and the Member States may legislate and adopt legally binding acts in that area. The Member States shall exercise their competence to the extent that the Union has not exercised its competence. The Member States shall again exercise their competence to the extent that the Union has decided to cease exercising its competence.’ (EC, 2012, p. 50).

(EU, national, regional, local) (Ahner, 2009). Sharing of experiences and inspiration between regions, and mutual cross-border cooperation are the essential elements of solidarity between EU member states. Namely cross-border cooperation, however, often faces a number of administrative, political, cultural and language barriers (to learn more see i.e. Perkmann, 2003; Perkmann, 2007; O’Dowd et al., 2013).

2.2 From common market to territorial cohesion The birth of the European Economic Community (EEC) in 1957 (at the same time, the European Atomic Energy Community (EAEC) was established) was associated with the support of economic integration of the founding states – Belgium, France, Italy, Germany, Luxembourg and the Netherlands. The EEC and EAEC continued the well-established cooperation between the states within the framework of the European Coal and Steel Community (ECSC), founded six years earlier in 1951. The efforts of the Western Europe to regulate nationalist state structures (in response to the war conflicts of the first half of the 20th century), create a common market and integrate European countries (in reaction to the growing threat of the socialist eastern bloc) resulted in economic and political cooperation and in the unification of European countries. In fact, the removal of tariff barriers and creation of a common market has polarized the territory of the member states into economically advanced regions and less developed regions. While a certain number of regions benefited from the barrier-free trade, some regions, however, were disadvantaged by their peripheral location, structural handicaps (e.g. old industrial cores negatively affected by deindustrialization processes) or unsuitable social and demographic structure. Ideas concerning possible solutions to unequal economic development began to emerge already in the early days of European integration. These considerations, reflected in the Spaak Report2 in 1956, have not met with much success and the issue of linkage between national policies and the EEC policies was ignored to a great extent. Nevertheless, the Treaty establishing the European Economic Community (EC, 1957) can be considered the first official document mentioning that the Community should support ‘a harmonious development of economic activities’ (EC, 1957, Article 2, p. 7) and ‘strengthen the unity of their economies and ensure their harmonious development by reducing the differences existing between the various regions and by mitigating the backwardness of the less favoured’ (EC, 1957, Preamble, paragraph 5, p. 8). In the context of continuing European integration, attention was paid to the removal of trade barriers. Disparities between regions were expected to be mitigated by market mechanisms and interventions from the Community were considered undesirable. A regional policy was a purely national issue. Regional inequalities were, however, relatively high

2 The Spaak Report was named after the former Belgian Foreign Minister Paul-Henri Spaak. The report proposed ways for the continuation of European economic integration and became the basis of the Treaty of Rome (Faludi, 2010). The authors of the Spaak report were aware of the new differences in the European area, resulting from the modification of economic ties across all member states. Thus, among other things, the report offered recommendations including the establishment and integration of a regional fund that would promote balanced regional development and interconnection of national and regional policies and future EEC policies (Faludi, 2009b).

(measured by economic indicators, mainly GDP per capita3). But only Italy – because of its higher level of inter-regional disparities (strongly affected by the backwardness of Mezzogiorno region) – advocated establishment of a supranational regional policy (König et al., 2009; Faludi, 2010). One of the first arrangements with the characteristic of regional policy is the foundation of the European Investment Bank (EIB), which was established in order to provide loans for development projects, and also the European Social Fund (ESF), which aimed at more effective provision of jobs and higher living standards. The EIB and ESF were part of the Treaty of Rome. But the ESF budget was rather negligible with limited impact on regional development (Blair, 2005). As late as 1968, Directorate-General for Regional Policies and Cohesion (DG XVI) (currently Directorate-General for Regional and Urban Policy (DG REGIO)) was established. Its main purpose was to improve the adverse situation of regions affected by the decline of heavy industry and to help rural regions lagging behind (Manzella and Mandez, 2009). The first measure focusing on levelling out differences between regions with a real budgetary nature was adopted at the Community level in the 1970s. At this time, regional policy began to be seen as the essential factor strengthening the Community. Establishing a common regional policy was partly related to existing common market, which was unable to mitigate differences between regions, and partly to desired reduction of economic disparities before the planned implementation of the single currency project (George, 1996). The forthcoming monetary integration became an impulse for the adoption of a regional policy because the adoption of the single currency expected national governments to stop using a number of national macroeconomic tools (such as decisions on inflation and state debt), which would be used to influence regional differences in economic growth (Fiala and Pitrová, 2010). The Thomson Report from 1973 became an important aspect influencing the implementation of regional policy. The report put regional development into the context of the common market and currency (Manzella and Mendez, 2009). Regional policy partly replaced national interventions by direct provisions and assistance from the Community. Supporting the poorest regions through financial resources provided at the EU level dates to 1975, when the European Regional Development Fund (ERDF) was established4 (Rauhut and Ludlow, 2013). With regard to the accession of new member states in the 1980s (Greece in 1981, Spain and Portugal in 1986), the Community adopted the Single European Act (SEA) aiming to reform the European institutions and accelerate the decision-making process allowing the implementation of the single market and its extension to new member states (Blair, 2005). Inter- regional differences (especially at the economic level) within the European Community were

3 George (1996) states, that in 1960s, GDP per capita was approximately four times higher in the richest regions than in the poorest ones. 4 The establishment of the ERDF is related, inter alia, to the enlargement of the EC in 1973 when three countries with less developed economies or larger disparities between regions joined the Community (Denmark, Ireland and the United Kingdom). Together with Italy, these countries became the proponents of regional policy within the Community (Loughlin, 1996). Mainly UK pressed the EC, demanding compensation for its high contribution to the Community budget and low income from the Common Agricultural Policy (CAP) resulting from its small- scale agriculture (Allen, 2005). The requested compensation consisted in support for outdated and declining industrial areas (Faludi, 2010). The newly acceding states and Italy created the so-called ‘cohesive coalition’ and in the end, gradually increasing the pressure on other states, managed to carry through the establishment of the fund (Fiala and Pitrová, 2010).

27 dramatically increased as a consequence of the accession of Mediterranean countries. The ratification of the SEA was crucially important for regional policy as economic and social cohesion was made one of the goals of the European Community5. The SEA added Title V called ‘Economic and social cohesion’ (EC, 1986) into the third part of the Treaty establishing the EEC. The newly added Article 130a includes the following statement (EC, 1986, p. 9): ‘In order to promote its overall harmonious development, the Community shall develop and pursue its actions leading to the strengthening of its economic and social cohesion. In particular the Community shall aim at reducing disparities between the various regions and the backwardness of the least-favoured regions.’ For the first time, the Treaty establishing the EEC provided information about structural funds – the ERDF, ESF and the European Agricultural Guidance and Guarantee Fund (EAGGF). The position of regional (structural) policy was strengthened in 1993 due to the emergence of a new political and economic structure called European Union (EU). Strengthening the economic and social cohesion has become one of the main EU activities. The first part of the Maastricht Treaty 'Common Provisions' specified that the new task of the Community is to support economic and social cohesion (EC, 1992). As a consequence, the Cohesion Fund (CF), which 'will provide Community financial contributions to projects in the fields of environment and trans-European network' (EC, 1992, p. 203), was established. The Maastricht Treaty also states that executive decisions related to the ERDF are adopted by the Council after consultation with the European Economic and Social Committee (EESC) and the newly established European Committee of the Regions (CoR)6 (EC, 1992). Investments in the regional policy funds were gradually increased, especially because of the needs of the newly acceding states in the pre-accession period before 2004. At that time, EU cohesion policy was mainly influenced by the Lisbon Strategy, which modified the institutional framework of the EU (with the expected enlargement of the EU between 2004 and 2007 contributing to increased regional inequalities)7. As a result, Lisbon Treaty extended the provisions regarding cohesion policy with explicit inclusion of territorial cohesion among the objectives of the EU.

5 The competence of the Community was extended to a number of new fields of interest such as the environment, social policy or science and technology policy. Blair (2005) argues, that the inclusion of a legal basis for environmental policy into the SEA represented a shift in the focus of Community’s attention towards policies traditionally associated with national governments and strengthened the legislative and regulatory role of the Commission within the whole Community. Thus, the Commission played a more prominent role in many international negotiations and the EEC has become a global actor operating internationally and predominantly addressing the economic and environmental issues and development cooperation. Changing European policies began to attract various stakeholders and lobbyists which created pressure on European Commission. 6 In the Maastricht Treaty, the CoR is defined as an advisory body ‘consisting of representatives of regional and local bodies’ (EC, 1992, p. 81). Until then, only the Committee on Regional Development (REGI) dealt with the issue of regional policy. The aim was therefore to incorporate representatives of individual regions into decision- making processes (Boháčková and Hrabánková, 2009). 7 While seven Romanian and Bulgarian regions reached less than a third of the EU average GDP per capita in 2007, 19 regions (including 11 capital city regions) achieved more than 50% above the EU average GDP in the same year (EC, 2010b).

2.3 Territorial cohesion as a part of EU cohesion policy In the official EU documents, the notion of ‘territorial cohesion’ was first used in the Amsterdam Treaty in relation to the meaning of Services of General Economic Interest (SGEI)8, which promote social and territorial cohesion (EC, 1997). The meaning of territorial cohesion was, however, very limited and no further specification was provided by the Amsterdam Treaty. The territorial dimension of cohesion obtained the legislative power with the ratification of the Lisbon Treaty during the European Council meeting in June 2007. The support of economic and social cohesion was extended with the territorial dimension. Article 2 of the Treaty on the European Union that covers EU goals, states that EU ‘shall promote economic, social and territorial cohesion, and solidarity among Member States’ (EC, 2007b, p. 11). As a part of the Consolidated versions of the Treaty on the European Union, the Article 4 classifies the economic, social and territorial cohesion as a shared competence between the EU and its member states (EC, 2012). Shared competence is an important basis for the collective promotion of territorial cohesion, but it must respect the principle of subsidiarity9. EU policies support the objective of territorial cohesion and the member states implement and coordinate their economic policies to achieve territorial cohesion. Title XVII (Title XVIII in the Consolidated versions of the Treaty on the European Union) was renamed to ‘Economic, social and territorial cohesion’ and Article 174, containing claims regarding the necessity of reducing differences between regions and the backwardness of the least-favoured areas, was complemented by the following statement (EC, 2012, p. 127): ‘Among the regions concerned, particular attention shall be paid to rural areas, areas affected by industrial transition, and regions which suffer from severe and permanent natural or demographic handicaps such as the northernmost regions with very low population density and island, cross-border and mountain regions.’ Although territorial cohesion became one of the EU goals, the Lisbon Treaty does not offer a more elaborate definition or at least general characterization. The annual Reports on economic, social and territorial cohesion pay attention to the state and perspectives of territorial development. The third report on economic and social cohesion highlights the meaning of territorial cohesion in rather general and complex nature (EC, 2004a, p. 27): ‘The concept of territorial cohesion extends beyond the notion of economic and social cohesion by both adding to this and reinforcing it. In policy terms, the objective is to help achieve a more balanced development by reducing existing disparities, avoiding territorial imbalances and by making both sectoral policies which have a spatial impact and regional policy more coherent. The concern is also to improve territorial integration and encourage cooperation between regions.’

8 To learn more, see Paper 5. 9 The subsidiarity principle means making decisions close to citizens and at the lowest possible level. Article 3b of the Treaty on the European Union states (EC, 2007b, p. 12): ‘Under the principle of subsidiarity, in areas which do not fall within its exclusive competence, the Union shall act only if and insofar as the objectives of the proposed action cannot be sufficiently achieved by the Member States, either at central level or at regional and local level, but can rather, by reason of the scale or effects of the proposed action, be better achieved at Union level.’

Based on the above definitions, territorial cohesion can be described as a very complex political concept whose primary objective is to reduce regional disparities through the support of disadvantaged regions. However, since its inception, territorial cohesion is closely tied to the issue of competitiveness, which is seen as an integral part of the successful economic development of the EU as a whole (EC, 1999; Sapir et al., 2004). The Territorial Agenda of the EU 2020 states that ‘the optimal balance of sustainability, competitiveness, and social cohesion can be realized through integrated territorial development’ (EU Ministers responsible for Spatial Development, 2011, p. 4). Moreover, global competitiveness of the EU and its regions is ranked among EU regional development priorities - their implementation should lead to greater territorial cohesion. Likewise, the recent Sixth report on economic, social and territorial cohesion, which focuses on investment for jobs and growth, explicitly states that ‘smart growth is needed to compete in the global market’ (EC, 2014, p. 54). Although the concept of territorial cohesion is characterized by a significant vagueness and complexity, the term has relatively quickly spread within the European institutions, and soon became a buzzword frequently used in the context of territorial development (Schön, 2005). Davoudi (2005) infers that such quick adoption of a rather vague term is related to its positive perspective. In other words, ambiguity becomes an advantage because the proponents of different approaches and beliefs may subscribe to territorial cohesion idea without committing to specific interpretation or application of the concept. Regarding the vagueness of territorial cohesion, Faludi (2005a) argues that territorial cohesion is a primarily political concept, whose purpose is to build a consensus. From the political point of view, strict definition of territorial cohesion would not be helpful in reaching agreements and compromises. Furthermore, territorial cohesion is valued thanks to its general and broad definition because certain fuzziness of territorial cohesion discourse is necessary to adapt its meaning to specific conditions of each region (Hamez, 2005; Camagni, 2007). Similarly, pragmatist theoretical planners support the idea of retaining the concept (for policymakers) fluid and fuzzy (Healey, 2009; Stein and Harper, 2012). Reactions to specific territorial development issues are then more innovative and reflect the unique contextual specificities of regions and the nature of the particular problem. Such understanding of the essence of normatively ‘defined’ territorial cohesion is presented in the document Community Strategic Guidelines on Cohesion (an indicative framework for preparations of operational programs in the period 2007–2013). What is significant is that ‘a different meaning should be given to territorial cohesion, linked to each Member State's history, culture or institutional situation’ (EC, 2006, p. 29). The importance of territorial cohesion should therefore not be universal and should be based on the specific characteristics of each region. Correspondingly, the Green Paper on Territorial Cohesion sees territorial cohesion as ‘a means of transforming diversity into an asset that contributes to sustainable development of the entire EU’ (EC, 2008, p. 3). Territorial cohesion, therefore, depends on a cooperation of all stakeholders across various sectors, as territorial issues relate to most of the sectoral policies. Excessive complexity and generality of territorial cohesion, however, is often criticized because the adoption of clearer objectives and more detailed specification of policy concepts contribute to a more coordinated territorial development and effective application of multi-level

30 governance (Servillo, 2010). According to Markusen (1999), the tasks of planners should be to remove the ambiguity of concepts in order to facilitate their operationalization and applicability. This view completely contrasts with the position of pragmatic planners campaigning for a looser definition of such concepts. Since spatial development strategies have an increasing impact on the EU regional policy as well as sectoral policies despite the non-binding character of EU spatial planning policy (see Section 2.4), a clearer definition of the meaning of spatial planning concepts should be European Parliament's priority not only for the EU institutions, but also for the wider scientific community (EC, 2008). Request for a more accurate definition of territorial cohesion has been raised by European Commissioner for Regional Policy Danuta Hübner, who claimed that ‘a clear and common understanding of this concept is necessary to communicate policy priorities to all stakeholders’ (Hübner, 2007). The following lines summarize the efforts to further specify territorial cohesion.

2.4 Specification of the concept of territorial cohesion

2.4.1 The role of ‘European spatial planning’ Although terminology related to the spatial organization of activities differs (‘town/country planning’ in the UK, ‘aménagement du territoire’ in France, ‘Raumplanung’ in Germany or ’planowanie przestrzenne’ in Poland), the most frequently used term at the EU level is ‘spatial planning’. In general, spatial planning is understood as planning of all the components of the environment, including the social one (Maier, 2004). The EU itself has no legal competence in the field of spatial planning and related activities come under the jurisdiction of individual member states (Faludi and Waterhout, 2002; Vanolo, 2010). Yet the European spatial planning discourse was quickly established and is now a significant part of EU policies. Already in the beginning of the European integration in the 1950s, some of the founding member states had experience with the planning of spatial development (e.g. France, Netherlands) and preferred to cope with the inter-regional differences in their own way (Kunzmann, 2006). Furthermore, planning traditions vary across European countries. In France, a relatively strong tool of the centralist government was ‘aménagement du territoire’, which aimed at balancing the huge economic gap between Paris and the underdeveloped rural west and south of the country. The state support of rural regions functioned as a counterweight to rapidly developing Paris. In Netherlands, ‘ruimteliike ordeningen’ is associated with important land resources management traditions (e.g. including the need to protect the area from possible floods) and represents the main activity of the public sector in terms of spatial development (Faludi, 2010). In Germany, in contrast with the French model, spatial planning is a matter of relatively autonomous federal states (Faludi, 2009b). The debate about the need and objectives of supranational regional planning in Europe (the term ‘spatial planning’ was not yet widely used) has been launched by a document of the Council of Europe from 1968 called Regional Planning a European Problem, which proposed regional planning as a feasible tool for development of individual countries as well as Europe as a whole (Kunzmann, 2006). At European level, regional planning became the subject of interest. In 1970, the tradition of annual meetings of ministers responsible for regional planning known under the acronym CEMAT (Conference Européenne de Ministers Responsable pour

L'aménagement du Territoire) was established. The conference addressed the issues of spatial planning at the supranational level (Faludi, 2010). Although the Council of Europe lost its interest in regional planning in the 1970s (the activities of the CEMAT within the Council of Europe, however, continued), the introduction of the ERDF strengthened the role of the European Parliament and the European Commission in this issue. The document European Charter of Regional/Spatial Planning, which was probably the first official document at the European level using the term ‘spatial planning’ and which established a set of fundamental objectives for territorial development in Europe, was adopted at the CEMAT conference in Spain in 1983 (Kunzmann, 2006). An important policy document Europe 2000, although not legally binding, was published by the Commission in 1991. The document focused on the importance of intangible investments and human resources, addressed the role of municipalities, regions and businesses in regional development, promoted creation of technological parks and an entire chapter was devoted to specific types of areas, such as islands and border regions or areas situated close to the external border of the Community (EC, 1991). Three years later, the document Europe 2000+ further deepened the institutional responsibility for the European territory (EC, 1994). The principal document of spatial development at the European level, the European Spatial Development Perspective (ESDP), was approved by the ministers responsible for spatial planning in Potsdam in 1999. Despite its non-binding character, formulation of a common view of EU countries on regional development was very important in order to increase the effectiveness of coordination of EU sectoral policies and cooperation among states, regions and cities (Faludi and Waterhout, 2002; Skokan, 2008; Faludi, 2009a; Vanolo, 2010). The fundamental objective of the ESDP is to achieve balanced and sustainable development by encouraging economic and social cohesion, preserving natural resources and cultural heritage, and aiming at the more balanced competitiveness of the European territory. Polycentric development of the EU territory10 and fair access to infrastructure and knowledge are the cornerstones of the ESDP objectives. The notion of territorial cohesion is not elaborated in the ESDP. The term itself is mentioned here only once in relation to harmonic, balanced and sustainable development (EC, 1999). ESDP terminology was, however, soon abandoned by the Commission. While the ESDP used terms like ‘spatial planning’ or ‘spatial development’, the later political and strategic documents replaced ‘space’ with ‘territory’ and focused on the support of ‘territorial cohesion’. Changes in terminology were associated especially with the negative attitude of some of the member states which did not consider spatial planning an EU competence. The transformation of terminology has been legitimized by the incorporation of territorial cohesion in the Lisbon Treaty and the growing emphasis on territorially integrated approach within all European and national policies (Schön, 2005). As a consequence, the Commission no longer used the term ‘spatial planning’ and the objectives developed within the ESDP framework became a part of the territorial cohesion policy (Faludi, 2009b). Thanks to Michel Barnier (already the

10 Supporting balanced polycentric urban systems and a new relationship between cities and rural areas became the key principles of the ESDP already at the meeting of ministers responsible for spatial planning in Leipzig in 1994 (ESPON, 2005).

Commissioner for Regional Policy), more attention to territorial cohesion was given in the Second report on economic and social cohesion from 2001, including a whole chapter dealing with territorial cohesion (Faludi, 2009b). It builds on the ESDP objectives and sees territorial cohesion as a path to more balanced development (EC, 2001). An informal meeting of ministers responsible for spatial planning together with the European Commission's DG Regio11 took an active part in preparing the ESDP and other documents focusing on territorial development (Territorial Agenda of the EU, Territorial Agenda of the EU 2020). Being the first such document, the ESDP not only significantly accentuated the issue of spatial planning but also influenced the formation and shape of interconnected programmes and initiatives oriented on European spatial planning (Waterhout, 2008). The INTERREG programme, designated to mitigate negative impact of borders, as spatial barriers, on territorial development, and the ESPON12 programme, established to provide policy makers at European, national and regional level with proper knowledge regarding territorial trends and potential policy impacts, are two of the most significant programmes (Kunzmann, 2006; Vanolo, 2010). One of the key tasks of the ESPON is to clarify the term territorial cohesion and to provide clearer and more understandable definition.

2.4.2 Multidimensionality of territorial cohesion Territorial cohesion, as understood by the EU authorities responsible for cohesion policy, is a complex concept that encompasses a wide range of activities and spheres of human lives. Complexity and broad scope make territorial cohesion a multidimensional concept. The meaning of territorial cohesion, as a strongly political concept, adjusts and adapts to current conditions within continuously transforming political objectives. When interpreting the concept, there are several schools of thought and proponents of different theories, which is then reflected in the attempts to define territorial cohesion. As already mentioned, the clarification of territorial cohesion is one of the main objectives of the ESPON (ESPON, 2014a). In most ESPON projects, definitions/dimensions of territorial cohesion are based on the ‘tree model’ recommended by the methodological guide Handbook on Constructing Composite Indicators provided by the Organisation for Economic Co-operation and Development/Joint Research Centre (OECD/JRC). The theoretical framework of the concept stems, in accordance with the methodology, from three steps: i) defining the concept; ii) determining the subgroups of the concept (essential dimensions); iii) identifying the selection criterion for indicators (OECD, 2008, p. 22, as quoted in Abrahams, 2014). Indicators are then compiled on the basis of data available for their measurement. According to the ideal process of the tree model, the concept of territorial cohesion can be broken down into several dimensions, which can be further quantitatively measured. The particular dimensions then characterize the basic features

11 Issues of territorial development and spatial planning are further supported particularly by DG Environment, the European Parliament (reports and resolutions related to spatial approach), the CoR and the EESC. 12 At first, the ESPON 2006 Programme, the European Spatial Planning Observation Network, was adopted. For the following planning period, the ESPON 2013 Programme, the European Observation Network for Territorial Development and Cohesion, was adopted (ESPON, 2014).

33 of the concept and can be relatively easily separated in terms of their fundamental meaning. Dimensions (grouped together) describe the exact meaning of the concept (see Figure 1).

Fig. 1. The tree model used to define the multidimensional phenomenon according to Abrahams (2014) and OECD (2008).

The most visible attempt to define territorial cohesion is evident in projects using methods of evaluation developed within the ESPON platform, which are the Territorial Impact Assessments (TIAs) and the European Territorial Cohesion Indicators (ETCIs)13. Delimitation of the basic territorial cohesion dimensions differs according to particular projects (see Table 1 based on the research by Abrahams, 2014) in spite of the fact that all projects operate with territorial cohesion as their primary objective. The five ESPON projects are supplemented with the priorities of spatial development which should lead to territorial cohesion and which are set by the Territorial Agenda of the EU and its updated version Territorial Agenda of the EU 2020. Similarly to the ESDP, the Territorial Agenda of the EU 2020 is based on a meeting of ministers responsible for spatial development (spatial development). In this sense, both Territorial Agendas are successors of the ESDP, and their main task is to strengthen territorial cohesion. The list of dimensions and the methods of their determination indicate that universal definition of territorial cohesion is very problematic. The request for clarification of the concept from EU political leaders has not yet been met, at least in terms of a simple and clear definition. Moreover, ignoring of specific features of particular regions regarding their spatial aspects, history, culture, socioeconomic and demographic situation is typical for a number of definitions. Despite the considerable diversity of basic dimensions of territorial cohesion, two key approaches (projects) that are essential for the measurement and evaluation of territorial cohesion may be identified: ESPON 3.2 TEQUILA (development of the TIA method) and INTERCO project (an attempt to synthesize diverse dimensions of territorial cohesion).

13 Since the results of projects in the new planning period ESPON 2020 are not yet fully known, the results of the ESPON 2006 and ESPON 2013 projects are used for the assessment of the multidimensionality of territorial cohesion.

Tab. 1. Definitions of territorial cohesion based on five ESPON projects and Territorial Agendas. Essential dimensions of territorial cohesion Project Method for definition (traits/themes/objectives/priorities) ESPON 3.2 . Territorial efficiency Traits identified through TEQUILA . Territorial quality policy analysis - TIA . Territorial identity (ESPON, 2006a)

Project 4.1.3 . Balanced distribution of population Traits identified through - ETCI . Sustainable structures policy analysis (ESPON, 2006b) TC = polycentricity: . Access to services . Avoiding negative externalities of excessive concentration of population, traffic, etc. . Avoiding excessive disparities in terms of income and wealth . A limited use of surfaces and environmental resources for human activities

TIP TAP . Territorial efficiency Traits identified through - TIA . Territorial quality policy analysis (ESPON, 2008) . Territorial identity

KITCASP . Economic competitiveness and Themes identified - ETCI innovation through workshops (ESPON, 2012c) . Balanced regional development and settlement infrastructure alignment . Social cohesion and quality of life . Sustainable development and environmental quality . Territorial co-operation and governance

INTERCO . Strong local economies ensuring global Objectives identified by - ETCI competitiveness merging tree and storyline (ESPON, 2012a) . Innovative territories models invented by the . Fair access to services, markets and project jobs . Inclusion and quality of life . Attractive regions of high ecological values and strong territorial capital . Integrated polycentric territorial development

Territorial Agenda EU . Polycentric development and Priorities identified (EU Ministers innovation through networking of city through recommendations responsible for Spatial regions and cities of the document ‘The Development, 2007) . New forms of partnership and territorial Territorial State and governance between rural and urban Perspectives of the areas European Union’ . Regional clusters of competition and innovation in Europe . Strength trans-European networks . Trans-European risk management including the impacts of climate change . Strength ecological structures and cultural resources

Territorial Agenda EU . Polycentric and balanced territorial Priorities identified 2020 development through recommendations (EU Ministers . Integrated development in cities, rural of the document ‘The responsible for Spatial and specific regions Territorial State and Development, 2011) . Territorial integration in cross-border Perspectives of the and transnational functional regions European Union – 2011 . Global competitiveness of the regions update’ based on strong local economies . Territorial connectivity for individuals, communities and enterprises . Managing and connecting ecological, landscape and cultural values of regions Source: Analyses of the ESPON projects according to Abrahams (2014)

Definition of territorial cohesion within the ESPON 3.2 TEQUILA project (Territorial Efficiency Quality Layered Identity Assessment) became the initial framework for most ESPON projects in the period 2000-2006 (as well as in other years). Territorial cohesion is here divided into three main dimensions (Camagni, 2006; ESPON, 2006a; Camagni, 2009). The first dimension is ‘territorial efficiency’, based on resource efficiency with respect to energy, land and natural resources, competitive economic structure and attractiveness of local territories, and internal and external accessibility. The second dimension is ‘territorial quality’, i.e. the quality of living, working environment, comparable living standards across territories and access to services of general interest (SGI) and knowledge. The third component is ‘territorial identity’ which is formed on the basis of territories’ social capital, the ability to develop a common vision for the future, local know-how and specific characteristics, productive expertise and competitive advantages of territories. Respecting the tree model, each dimension is decomposed into several sub-dimensions (see Figure 2). This was one of the first definitions of territorial cohesion as an overarching political goal.

Fig. 2. The TEQUILA model of territorial cohesion. Source: ESPON (2006a), Abrahams (2014)

On the contrary, the INTERCO project (Indicators of Territorial Cohesion) was one of the latest attempts (within the ESPON platform) to explain the importance of territorial cohesion. The aim of the project was to ‘develop indicators and indices that can be used to measure territorial cohesion’ (ESPON, 2012a, p. 4). First, however, territorial cohesion had to be defined. Based on existing documents (ESDP, cohesion reports, Territorial Agenda EU, Green Paper on Territorial Cohesion, previous ESPON projects) and intensive discussions on territorial cohesion, the project team was aware that simple definition of territorial cohesion does not exist. Relevant topics related to territorial cohesion (viewed from political and spatial planning perspective) became a platform for the establishment of main territorial cohesion objectives. The INTERCO project used not only political documents and scientific literature but also participatory interaction with stakeholders (ESPON, 2012a). The characteristic of territorial cohesion was also dependent on the possibility to practically evaluate its specific aspects/dimensions (criterion of usability) at various spatial scales. As the only one of the selected ESPON projects dealing with territorial cohesion, the INTERCO project focused on statements made by the involved stakeholders. Their engagement in discussion enriched thinking about territorial cohesion and was reflected in the complexity of the resulting definition (see Subsection 2.5.2). The definition consists of six basic dimensions (goals), which are specified by the set of indicators used for their measurement (see Figure 3).

Fig. 3. The INTERCO model of territorial cohesion. Source: ESPON (2012a)

The efforts to operationalize territorial cohesion through dimensions and sub-dimensions suitable for evaluation of territorial cohesion (the TIA and ETCI) are typical also for the TEQUILA and INTERCO projects. The comparison of the two ‘definitions’ of territorial cohesion brings, again, a number of dissimilarities and in the end, the concept of territorial cohesion becomes even fuzzier and more ambiguous. Defining territorial cohesion through various partial aspects is, in the case of multiple approaches (i.e. projects), rather confusing. The question, therefore, arises whether such a method is appropriate for clarification of such a complex concept – which territorial cohesion definitely is. Moreover, it is evident that certain meanings of territorial cohesion are already operationalized and applied in practice in some of the member states, just under another specification and under differently defined spatial policies (Faludi, 2004; Abrahams, 2014). Attempts to define territorial cohesion through dimensions and indicators are also complicated by different interpretations of the concept that may be mutually related or contradict each other. The result is several simultaneously existing narratives conceptualizing territorial cohesion differently. The following Section views territorial cohesion as a discourse within which the basic features and interpretation are being deciphered.

2.5 Territorial cohesion as a discourse Competing interpretations of territorial cohesion are based on different traditions of spatial planning in particular EU countries, which stem from their specific historical development, and

38 also on concurrently proclaimed goals of cohesion policy. The issue of exercising power is closely linked to assigning meanings to territorial cohesion. First of all, it is useful to get acquainted with power relations and settings that lay behind the process of adopting the idea of territorial cohesion. After that, the key narratives (referred to as ‘storylines’) serving as a specific interpretation of territorial cohesion will be provided. Subsequently, the focus will be put on two crucial storylines - cohesion and competitiveness - that have resonated throughout the discussions on territorial cohesion in recent years and which pervade all attempts to define territorial cohesion. In this context, the issue of polycentric development seems to be the key one.

2.5.1 The roots of territorial cohesion Initially, the EU single market was based on a neoclassical economic model assuming that competition within larger markets will generate sufficient profit to ensure the development of poorer regions. In accordance with the convergence theory, which presupposes a tendency toward mitigating inter-regional differences, there were hardly any market interventions14 (Blažek, 2008). With the EU expansion, however, inter-regional differences started to grow, which was the main stimulus for the emergence of regional policy with interventionist elements and subsequent inclusion of territorial cohesion among the official EU goals. The gradual formation of the non-binding European spatial planning policy was the main incentive for accentuating the spatial approach and opening the discussion on territorial cohesion. As an unofficial part of the cohesion policy, European spatial planning is strongly linked with a number of interrelated programs and initiatives, in which representatives of various member states participate. The logical consequence is that different traditions of spatial planning in European countries are reflected in the conceptualization of territorial cohesion. Territorial cohesion principles are strongly inspired by French policy ‘aménagement du territorie’, which is the French equivalent of spatial planning and is primarily based on the regional economic approach (Faludi, 2009b). The main objective of the regional economic approach is the development of regions stemming from a generally accepted strategic framework which is formulated by the central authority and which provides funding and coordinates the development of particular territories (Faludi, 2004). Dupuy (2000) believes that the ambitions to centrally coordinate the regional policy are rooted in the interest of the French to maintain their national unity15. The strongly centrally controlled French regional policy mainly responded to the ongoing economic changes caused by the loss of the French colonies and the loosening of trade barriers within the EEC in the 1960s (Burnham, 1999, as quoted in Faludi, 2004). The economic dominance of growing Paris (in terms of population and at the

14 The basic division of regional development theories (Blažek, 2008) divides approaches to regional development on the basis of the development trend. Regarding the convergence theory, tendencies to reduce inter-regional disparities prevail. From the perspective of the divergence theory, growth of regional differences over the course of time is a typical feature. The main difference between the theories is the various significance ascribed to the differentiation (e.g. cumulative, concentrative, selective) and equalizing mechanisms. Nevertheless, Blažek (2008) argues that research of regional development trends is influenced by many factors (e.g. selection of indicators, length of the period under review, hierarchy of territorial units) and should not be overestimated. 15 The national French institution DATAR (Délegation à l´aménagement du territoire et à l´action régionale), tasked with coordinating the activities of various ministries regarding regional development, managing regional funds, and reducing overall economic differences between regions, was established in 1960s.

39 expense of most of the other French regions), and consequently the growing differences between the capital and especially rural regions, was another important feature contributing to the emergence of the territorial cohesion discourse within the framework of the French regional policy. ‘Aménagement du territorie’ involves public activities related to the distribution of people, actions, and physical structures in space that is based on the imagination of balanced development with regard to the geographical and social situation in a particular region (Dupuy, 2000). It is, in other words, a strategic spatial framework designated to allow eventual interventions by the public administration (Faludi, 2009b). Enforcement of the French regional development model was associated with the relatively important role of France in the context of regional policy and European integration. France accepted a certain weakening of its autonomy under the condition of achieving a more dominant position within the Community (Eising and Kohler-Koch, 1999). According to some experts (e.g. Levy, 1997, as quoted in Faludi, 2004; Bailly 2001), the DATAR started to promote the concept of ‘aménagement du territorie’ at the EU level, resulting in the establishment of structural funds, which were based on the DATAR model (redistribution of funds in order to reduce inter-regional inequalities). The fundamental principles of territorial cohesion, stemming from ‘aménagement du territorie’ are not typical only for France, where this tradition of spatial planning originated, but in some ways, they were applied by central planning authorities in socialist regimes. In this context, it is obvious that the formation of spatial planning discourses was carried out in different political and economic systems based on contrasting ideologies. While in Western Europe democratic systems with the market economy were formed, the political system of the Eastern bloc relied heavily on central state control with little influence of local actors. Despite different political and ideological frameworks, similar principles can be identified in the practices of spatial planning, which was in both cases based on reducing regional inequalities and supporting the regions lagging behind, (Malík, 1976; EC, 2008). The empirical part of the dissertation (see Paper 1) aims at finding parallels between the two doctrines of spatial planning (and thus answering the particular research questions). In contrast with the regional economic approach, the comprehensive integrated approach is characterized by the existence of development plans at different hierarchical levels (Faludi, 2004). Systematic application of plans should lead to coordination of the public sector activities with regard to effective territorial development. This approach requires planning institutions and mechanisms and considerable political commitment. While the regional economic approach focuses on the localization of economic activities in order to initiate local/regional development of the disadvantaged areas, the comprehensive integrated approach seeks to balance the development requirements and the carrying capacity of the area, i.e. land. The integrated approach, therefore, incorporates the principle of sustainability and gains significance especially in densely populated areas (e.g. the Ruhr, Benelux), with higher pressures on land use (Faludi, 2004, 2005a), and is typical for the Netherlands or Germany. The attitude of the integrated approach towards economic development is changing and in some cases (e.g. high unemployment rates in some regions) regulatory measures are replaced by economically oriented interventions reflecting the ‘need’ for economic competitiveness. Similarly, the regional economic approach, in the context of the increasing importance of sustainable development, adopts the mechanisms typical of the integrated approach.

The current EU regional policy includes measures characterized by a variety of approaches to regional development. Although the concept of territorial cohesion is promoted at European level, strong endogenous development orientation, i.e. the use of local and regional potential, is supported in order to reach a certain level of cohesion. Blažek and Uhlíř (2011) notice a gradual enlargement of the thematic scope of regional policy (from the support of infrastructure and business to development of human resources) and especially the incorporation of the territorial dimension within sectoral policies, which often have a greater regional impact than the actual regional policy. In this context, the attitude towards territorial cohesion differentiates, depending on the tradition and form of spatial planning in individual member states. As soon as in the late 1980s, the Netherlands together with France proposed a spatial framework for managing the structural funds (which resulted in the ESDP), but they did not consider the question of power configuration as the ‘aménagement du territorie’ policy did not address competencies at all. Germany, where the key power regarding regional development falls within the remit of individual federal states and spatial planning is strictly separated from regional policy, pointed to the unresolved competence issue. Germany did not belong to supporters of the functioning of regional policy according to the Commission guidelines and still shows reservations regarding the issue of accepting territorial cohesion (Faludi, 2009b). Similar scepticism comes from the UK, which claims that support of disadvantaged regions with specific geographical features is a matter of each member state and that each state supports territorial cohesion individually through applying specific measures, thereby creating European added value. Contrarily, the role of the Commission in supporting transnational cooperation is supported mainly by the greatest beneficiaries of the EU budget16 (including the Czech Republic).

2.5.2 Territorial cohesion ‘storylines’ The difference in perception between the member states and the EU concerning the importance and relevance of territorial cohesion is reflected in the difficult interpretation of the term. The result is a number of possible interpretations of territorial cohesion, referred to as ‘storylines’. Unlike the tree model, the storylines concept is sceptical about using clearly defined basic dimensions which together constitute the meaning of such umbrella concepts as territorial cohesion (Hajer, 1993, 1995). On the contrary, the individual storylines are more or less mutually competing interpretations. The concept of storylines itself is closely connected to discourse analysis theory. As Hajer (2000, p. 137) states, ‘discourses should not be seen simply as “debates” or “discussions”’. The true essence of discourse lies in the fact that it is ‘an ensemble of ideas, concepts, and categorizations that are produced, reproduced and transformed in a particular set of practices and through which meaning is given to physical and social realities and which permeates regional, national and supranational policy making circuits’

16 Although the decentralization trend is evident in the decision-making processes and competences have been transferred to the lower spatial administrative level in recent years in Europe, a range of competences are transmitted to the supranational EU institutions in line with the continuing European integration process. Thus, there are ongoing negotiations at several levels – European, national, regional, local – i.e. the process of multi- level governance. Competences of individual institutions are a part of a complex system of decision-making processes, which vary across member states. The position of member states towards competences of the Commission, in the context of territorial cohesion, also differs depending on the basis of specific systems of governance (for more on multi-level governance see e.g. Servillo, 2010).

(Hajer, 1995, p. 44). According to Waterhout (2008), Hajer’s interpretation of discourse is derived from Foucault’s abstract understanding of discourse (Foucault, 1969), but at the same time, it is modified in order to facilitate its application to specific political events. In practice, this transformation is accompanied by an opportunity to identify ‘discourse coalition’ and the set of ‘storylines’. Hajer (2000, p. 139) claims that discourse coalitions should be viewed as ‘a variety of actors that do not necessarily meet but through their utterances reinforce a particular way of talking that is reproduced via an identifiable set of storylines and discursive practices in a given policy domain’. Regarding such theoretical framework, Waterhout (2007) perceives that the given policy domain is territorial cohesion, which is reproduced by partial storylines. Storylines are then defined as ‘(crisp) generative statements that bring together previously unrelated elements of discourse and thus allow for new understandings and create new meanings’ (Hajer, 2000, p. 140). One of the significant features of storylines is their metaphorical character, which allows for a diversity of interpretations. This is the main reason for the emergence of discourse coalitions bringing together different sets of actors ‘that do not necessarily share a particular set of belief or normative commitment’ (Hajer, 2000, p. 140) but are bound together by multi- interpretable storylines. The following lines summarize the key territorial cohesion storylines (Table 2 provides an overview). Based on policy analysis, Waterhout (2007) defines four storylines that he labels as the seedbed for territorial cohesion policy. These are i) ‘Europe in balance’; ii) ‘Competitive Europe’; iii) ‘Coherent European policy’; iv) ‘Green and clean Europe’. The first storyline ‘Europe in balance’ currently dominates the territorial cohesion discourse. It concerns especially marginal areas that are characterized by a greater degree of peripherality and spatial exclusion. The goals of fair access to SGI and balanced polycentric development are supported. Conversely, all the other storylines are geographically focused on the entire EU area. In contrast to the ‘Europe in balance’ storyline the ‘Competitive Europe’ is targeted at Europe's global and regional competitiveness. Advocates of this approach argue that territorial development cannot be achieved via redistributive tools but through added value of regions which benefit from their unique territorial capital. In this case, polycentric development is mentioned in relation to the need to create strong metropolitan centres that would compete with the current EU economic core (EC, 1999). The aim of the ‘Coherent European policy’ storyline is to achieve horizontal coherence and thus create a spatial framework that would coordinate and integrate EU sectoral policies which are often characterized by unintended negative spatial impacts. The storyline ‘Green and clean Europe’ is tied to the goals of sustainable development and effective environmental management. Very similar structuring is provided by Evers et al. (2009), which through literature overview identifies five crucial interpretations (storylines) of territorial cohesion discourse. The aspect of balanced development is a part of the interpretation ‘Territorial cohesion and socio- economic convergence’, which aims at the support of spatially disadvantaged areas in order to reduce inter-regional disparities. The second interpretation called ‘Territorial cohesion and economic competitiveness’ emphasizes the effective use of territorial potential, which is supposed to lead to the EU’s competitiveness on the global market. The interpretation ‘Territorial cohesion and rural perspective’ is closely linked with the issue of depopulation of

42 rural areas facing inadequate supply of public services and declining income in the agricultural sector. The fourth interpretation ‘Territorial cohesion and spatial planning’ focuses on the proper role of territorial management discussed in recent years. Spatial planning at European, national and regional level should be a prerequisite for tackling common European challenges, such as the unbalanced regional development, adaptation to climate change, urban deprivation and sprawl. The last interpretation ‘Territorial cohesion as policy coordination’ understands territorial cohesion as a ‘key for resolving conflicts and creating synergy between sectors and tiers of government’ (Evers et al., 2009, p. 11). Diverse sectors of the EU policies should reflect the spatial effects of their activities. Territorial cohesion storylines defined by the INTERCO project were created on the basis of political objectives and their final shape was largely modified by testimonies and perspectives on territorial cohesion provided by different groups of stakeholders. The result was five storylines (ESPON, 2012a), whose importance is summarized by Gløersen and Böhme (2011). The storyline ‘Smart growth in a competitive and polycentric Europe’ is linked with the priorities of the strategic document Europe 2020, mainly greater economic growth and better European competitiveness. The main goal of the storyline ‘Inclusive, balanced development and fair access to services’ is to fulfil the wishes that people should not be disadvantaged by where they live or work (EC, 2004a). This should be ensured by at least a minimum limit of provided services across the EU territory. The storyline ‘Local development conditions and geographical specificities’ puts emphasis on local development conditions and comparative advantage of particular places. At the same time, it draws attention to sparsely populated areas (including mountainous areas and islands), which suffer from the higher vulnerability of ecosystems, and limited availability of technical infrastructure and services. Interconnecting human activities to the physical environment is the main issue included in the storyline ‘Environmental dimension and sustainable development’. The storyline ‘Governance, coordination of policies and territorial impacts’ sees the role of territorial cohesion in the improvement of vertical and horizontal coordination of the EU sectoral policies. All approaches that identify storylines are characterized by mutual overlays. This means that some goals or priorities set by individual interpretations may also be symptomatic for interpretation within the frame of a different storyline (a typical example is the concept of polycentric development occurring in the interpretation of balanced spatial development as well as the interpretation emphasizing global competitiveness). This relates to the overlap of the actors within the discourse coalition typical for a particular storyline. In other words, certain actors (states, EU institutions) can support multiple storylines (see Waterhout, 2007). And although individual storylines, defined by various authors, may differ in terms of content, the original delimitation of storylines carried out by Waterhout (2007) seems to be very relevant also for other cases (see the main traits of storylines in Table 2). The issue of coordination of the various sectoral policies (‘P’) has been present in debates on territorial cohesion since the beginning, as well as the issues of sustainability and environmental friendliness (‘E’). Both of these storylines are also included in other interpretations (at least minimally). The main tension is observed between the fundamental interpretation of territorial cohesion as balanced Europe (‘B’) and the interpretation emphasizing competitive Europe and its regions (‘C’).

Tab. 2. Storylines of territorial cohesion. Source and method The main for identification of Storylines of territorial cohesion trait of storylines storylines * Waterhout (2007) . Europe in balance B - policy analysis . Competitive Europe C . Coherent European policy P . Green and clean Europe E Evers et al. (2009) . Socio-economic convergence B - policy analysis . Economic competitiveness C . Rural perspective B . Spatial planning P/E . Policy coordination P INTERCO . Smart growth in a competitive and polycentric Europe C - policy analysis and . Inclusive, balanced development and fair access to B workshops with services stakeholders . Local development conditions and geographical B/E (ESPON, 2012a) specificities . Environmental dimension and sustainable E development . Governance, coordination of policies and territorial P impacts * B = balance; C = competitiveness; P = policy; E = environment Source: Waterhout (2007), Evers et al. (2009), ESPON (2012a); author’s processing

2.5.3 Issues of cohesion and competitiveness Despite a number of potential territorial cohesion storylines, mentioned above, dual interpretation is specific for the conceptualization of territorial cohesion (Ache et al., 2008; Farrugia and Galina, 2008; Vanolo, 2010): a territorial cohesion narrative emphasizing harmonic and balanced development or a narrative accentuating support of competitiveness and economic growth. The objective of territorial cohesion covering aspects of social solidarity and welfare state may seem like a contradiction to the principles of neo-liberalism operating on the basis of the free market and fully utilizing competing mechanisms in today’s globalized environment. However, within the framework of European policies, territorial cohesion and competitiveness are relatively strongly connected, despite the fact that conceptualizing the relationship of both goals, expressed by two storylines, is not clear (Vanolo, 2010). Interpretation of territorial cohesion promoting balanced Europe is strongly linked with the French tradition of ‘aménagement du territorie’ and receives support from many member states (particularly those with the presence of backward regions and thus benefiting from structural funds). It is characterized by a strong support of cohesion objectives and redistribution of financial resources in order to reduce regional disparities. This view stems from the perception of globalization and European liberal economies as a threat to regional and spatial development, meaning that only the strongest regions and localities will be favoured, and other peripheral locations will further socially and economically decline (Robert, 2007).

Territorial cohesion is perceived from the traditional perspective of economic and social cohesion, which aims at moderating inter-regional inequalities. Proponents of this approach are European spatial planners responsible for creating the ESDP and lobbyists supporting the idea of equitable access to the SGEI (Faludi and Waterhout, 2002). Especially the Mediterranean countries with a higher degree of inter-regional disparities are among the major proponents. According to them, balanced territorial development is the prerequisite for the subsequent competitiveness of Europe as a whole. This is related primarily to polycentric development, which includes aspects of cohesion and harmoniously balanced development (EC, 1999; COR, 2005). In the context of achieving more territorially cohesive Europe, access to the SGEI is accentuated, especially their inadequate provision in remote regions (EC, 2004b). Territorial cohesion is associated with the development of the peripheral and most disadvantaged regions, which are characterized by worse accessibility and lower population density. Besides southern European states and Belgium or Luxembourg, better accessibility of SGEI is supported by associations such as the Assembly of European Regions (AER), the Conference of Peripheral and Maritime Regions of Europe (CPMR), and the CoR (Waterhout, 2008). On the other side, states supporting the storyline ‘Competitive Europe’ understand globalization as a phenomenon that should be used to European profit and its global competitiveness. Specifically, investments should be made into the most developed urban cores, European territorial structure should be optimized and the unique territorial capital of European territories should be exploited. The aim is not only global competitiveness of the whole Europe but also the competitiveness of individual regions. During the work on the ESDP, it became clear that the proponents of this approach are predominantly countries from north- western Europe (especially the Netherlands), which sees the development of Europe in supporting of existing core economic clusters. As pointed out by Zonneveld (2000), the role of Netherlands was crucial for shaping this storyline. Unlike traditional division of Europe into the core and the periphery, for the first time, the global position of Europe as an integrated system of urban networks that covers the area outside the traditionally perceived economic core (the so-called Pentagon area) and is the driving force of competitiveness was introduced. Waterhout (2008, p. 108) further adds that ‘the clearest example, though, of competitiveness being taken into account is the concept of global economic integration zones’. From the perspective of polycentric development, these zones should form a counterweight to the European economic core in order to ensure more balanced regional development. Polycentric development is understood both in the sense of cohesion and competitiveness. This is supported by the statement that ‘competitive and polycentric Europe’ needs to build ‘on territorial potentials, on the support of smart growth and on the connectivity of Europe's economic centres in a polycentric system of cooperation and integration’…then ‘balanced and harmonious territorial development would subsequently be achieved on the basis of diffusion processes’ (Gløersen and Böhme, 2011). Ensuring competitiveness precedes the subsequently formed territorial cohesion, as demonstrated by the recent Sixth cohesion report claiming that investment in regional competitiveness improves people's lives and is the basis for growth in less developed regions (EC, 2014).

In order to facilitate the interconnection of territorial cohesion interpretations (especially balance and competitiveness), the concept of polycentric development gradually gained popularity (Davoudi, 2003; Zonneveld et al., 2005). Polycentric development, supposedly able to bridge the cohesion and competitiveness dichotomy, is rather a result of political debates. Consequently, from the analytical perspective, polycentricity itself has rather unclear and ambiguous meaning (Krätke, 2001). Farrugia and Galina, who are sceptical about the role of polycentricity in achieving competitiveness, state that ‘if the path towards the competitiveness obsession continues, European spatial development future scenarios are less likely to be polycentric’ (Farrugia and Galina, 2008, p. 4). They refer to ESPON (2007) study, which offers two basic scenarios for the future spatial development of the EU (see Appendices 1 and 2). The scenario oriented to Europe's competitiveness assumes economic growth, especially in central regions, resulting in further deepening of regional disparities and environmental and social risks. At the national level, polarization of economic development will intensify due to the dynamic growth of major cities. The scenario of European cohesion has a positive impact on demographic trends, social and cultural integration and less negative impact on rural regions. Economic performance will not reach such a level as in the previous scenario. At the national level, the role of alternative metropolitan areas (others than major cities) will be strengthened. At the regional level, medium-sized cities will play the key role in local and regional development.

2.5.4 Polycentric development One of the main objectives and a key element of territorial cohesion is to promote polycentric and balanced spatial development (EC, 1999; EU Ministers Responsible for Spatial Development, 2007, 2011). The crucial goal of polycentric development is to spread economic and economically relevant functions among a larger number of urban centres with regard to balanced development across territory (Zonneveld et al., 2005). At the European level, cooperation of the most developed cities and regions should create added value and centres would positively contribute to the development of their wider regions (EU Ministers Responsible for Spatial Development, 2011). The support of polycentricity is focused on cities and regions that through networking and linkages with other cities can start their economic growth. The existence of urban networks is expected not only within the EU borders (or Europe) but also beyond them, where the linkages to cross-border territories and markets are an essential part of the EU territorial development (see Appendix 3). The territorial Agenda of the EU 2020 (EU Ministers responsible for Spatial Development, 2011) also states that polycentric development is desirable at the intra-national level, where is the need to avoid excessive territorial polarization between metropolitan areas and smaller regional centres and rural regions. The reducing polarization of economic performance should thus lead to avoidance of large regional differences. Territorial cohesion policy, therefore, supports polycentricity not only at European but also at the national and regional level. The nature of polycentric development, however, differs at various spatial scales and geographical contexts (Hall, 2002). At the global level, polycentricity points to the development of alternative global centres in terms of power (London and partly Paris may be considered so- called ‘global cities’ within the European territory). According to ESDP, polycentricity at

European level means primarily the support of centres outside the so-called Pentagon17 (the most economically advanced region in Europe, delimitated by London, Paris, Milan, Munich and Hamburg), i.e. mostly the capitals of other European countries (EC, 1999). Although EU belongs to the most economically advanced regions of the world, its internal structure is characterized by relatively large inter-regional disparities. The aim is to create more globally competitive and economically stronger areas that would ensure the future of the European economy (Davoudi, 2003). Globally and economically oriented approach is applied despite considerable environmental and social problems the economic core area of the EU is facing (Jenses and Richardson, 2001). Approach to development in terms of utilization of the inner potential of individual regions is represented by a model called ‘Bunch of Grapes’ (Kunzmann and Wegener, 1991, as quoted in Faludi, 2005b). Unlike the monocentric perception of Europe, it provides a polycentric view, where the development of regions involves exploiting the opportunities and advantages of each region. In line with territorial and social cohesion, regions that are lagging behind should be supported. At European level, polycentricity is understood as a principle ensuring competitiveness on the world market on the one hand and remodelling existing regional disparities into a more territorially and socially cohesive form on the other (Davoudi, 2003). The strategy of EU spatial planning policy includes also supporting polycentricity at the national and regional levels. At the national level, polycentric development is associated with balancing the dominant (usually capital) city through the support of smaller regional centres. In this context, polycentric development should aim at balanced spatial organization of human activities, which is expressed in the relationship between the core city and its hinterland within the urban region, interrelations between city regions of varying sizes and importance, and interactions of urban regions with areas situated behind their borders (Sýkora et al., 2009). Within the INTERREG IIIB program, the project RePUS (2007), focusing on regional polycentric systems in Central and Eastern Europe, emphasized the bottom-up perspective – based on daily urban systems, where the everyday life of people usually occur. Local urban systems and networks of regional centres have the potential to counterbalance the dominance of the major cities. Promoting competitiveness of major cities and metropolitan areas should, therefore, be accompanied by strengthening of the role of other large and medium sized cities which are an important element for regional development. Importance is also attached to smaller cities that are centres of employment and services at the local level (particularly in rural and peripheral regions), and thus ensure social and economic cohesion at the lowest level (RePUS, 2007). Essentially, promoting polycentricity on several spatial scales seems to be a contradictory step (Hall and Pain, 2006). At European level, polycentricity means strengthening the role of urban centres outside the most economically developed and densely populated area stretching from southeast England towards the south across France and Benelux to northern

17 Imaginations of the European economic core use different metaphors, while there are conflicting opinions on particular approaches. In addition to the ‘Pentagon’ area, the traditionally defined core region is called the ‘Blue Banana’ and includes southeast England, Benelux, northern France, midwestern and southwestern Germany, Switzerland and northern Italy (Brunet, 1989, as quoted in Davoudi, 2003); or the ‘Red Octopus’ with its core in Germany and spreading into surrounded areas comprising the largest European agglomerations (Van der Meer, 1998).

Italy. Thus, it is primarily the support of capitals in Central and Eastern Europe, whose potential economic growth accentuates, even more, the monocentric structure of national urban systems (the case of Hungary, the Czech Republic, Slovakia and other countries with a dominant position of a major metropolis). Basically, the promotion of polycentric urban systems at European level can lead to monocentric urban systems at the national level. A similar ambivalence can be observed at other spatial levels. Polycentricity is therefore strongly dependent on the scale. Polycentricity is based on two basic processes (Salone, 2004, as quoted in Farrugia and Galina, 2008). Polycentric development as a ‘structural’ process (morphological, economic, functional) stems from the internal spatial dynamics of a particular area (functional linkages within the urban network) and the distribution of urban centres (Kloosterman and Musterd, 2001; Parr, 2004). Polycentric development in the sense of an ‘institutional’ process refers to institutional integration and ‘multi-level polycentric governance’ (Hooghe, 1996; McGinnis, 1999; Finka et al., 2015). According to Farrugia and Galina (2008), polycentric development is characterized by three main components: i) a functional division of labour (as a structural process); ii) economic and institutional integration (as a structural/institutional process); iii) political cooperation (as an institutional/political process). The political precondition of polycentricity is its expected ability to generate growth and equilibrium. Nevertheless, depending on the choice between the neoliberal economic model (promotion of competitiveness) and inclusive policy (supporting cohesion), structural and institutional processes may change when the responsible political elites can selectively activate the capacities and competencies of a particular territory (Farrugia and Galina, 2008). Although, in political debates, polycentricity is seen as a bridging concept linking the competing interpretations of cohesion and competitiveness (normative aspect), the impact of polycentric development on territories remains, in fact, ambiguous, especially influenced by the geographical scale and the nature of processes through which polycentricity is evaluated (e.g. economic, social, cultural cooperation). When examining the meaning of polycentricity, it is, therefore, essential to consider its three main features: i) polycentricity as a path to cohesion or as a path to competitiveness; ii) the nature of the processes, polycentricity is based on; iii) geographical scale.

3 NARROWING DOWN TERRITORIAL COHESION COMPLEXITY

With respect to the formulation and purpose of research questions, the complexity of territorial cohesion has to be reduced. Thus, the empirical part of this dissertation works mainly with the meaning of territorial cohesion as a concept aiming at the balanced spatial development of a particular territory. Under this interpretation, the analysis focuses on the concept of polycentricity and the issue of equal access to SGI. The key spatial level is regional scale.

3.1 Territorial cohesion as a path to balanced development The interpretation of territorial cohesion through basic narratives brought fundamental knowledge about the two main competing semantic lines of territorial cohesion and competitiveness, which are characteristic for the current debate within the European spatial planning platform. Although the interpretation of competitiveness is still present in the territorial cohesion discourse (EC, 2010a, 2014), the empirical part of the dissertation uses predominantly the conceptualisation of territorial cohesion from the perspective of the balanced development of Europe and its regions for the following reasons: i) The term ‘cohesion’ is included in the name of the concept (competitiveness not) and the aspect of ‘better balance’ is contained in a single official document, the Lisbon Treaty, incorporating territorial cohesion among the common objectives of the EU and claiming that ‘in particular, the Union shall aim at reducing disparities between the levels of development of the various regions and the backwardness of the least-favoured regions’ (EC, 2012, p. 127). ii) For the first time, territorial cohesion was officially used in the Amsterdam Treaty in connection with SGEI (EC, 1997). Since then, accessibility of these services has been supported mainly by lobby groups supporting the interpretation of territorial cohesion as contributing to a reduction of inter- regional disparities (Waterhout, 2008). iii) Since territorial cohesion is a part of EU cohesion policy, which has very strong and relatively stable political support, territorial cohesion narrative emphasizing harmonious and balanced development has many supporters as well (Waterhout, 2007). Strong embeddedness of cohesion policy within the framework of European policies allows for easier acceptance of territorial cohesion as a way towards balanced development. iv) It is this interpretation that shows similar aspects to social-equalizing principles of the socialist central political apparatus. In particular, the principles of socialist urban planning exhibit similar traits, such as including the territorial cohesion objective in a narrative accentuating balanced development. This will allow comparing both doctrines of spatial planning. v) Despite the interpretation of balanced Europe within the territorial cohesion discourse, it has become increasingly often confronted with the principle of competitiveness in recent years (Doucet, 2006; Vanolo, 2010) and has been criticized for its lack of efficiency (Waterhout, 2007). Above all, the

relationship between equitable access to jobs and services (EC, 2003) and regional specificities, whose importance is also crucial for territorial cohesion is unclear (EC, 2008). In the case that removing territorial specificities is not desirable, but rather they are seen as potential advantages of development, application of the principle of equality and balanced regional development may seem as suppression of the unique characteristics of a particular territory.

3.2 Polycentricity and access to SGI Territorial cohesion discourse can be analysed by means of specific meanings and concepts that are crucial for its operationalization. In this context, the dissertation will focus particularly on the concept of polycentricity and access to SGI, mainly for the following reasons: i) The central theme of territorial cohesion is polycentricity (EC, 1999; Davoudi, 2003; EU Ministers responsible for Spatial Development, 2011). ii) Polycentric development is associated with fair access to SGI and jobs, i.e. the basic precondition for territorial cohesion established by the Amsterdam Treaty (EC, 1997). iii) Despite the considerable popularity of polycentricity within the framework of European policies, it's still a relatively vague term (Hague and Kirk, 2003; Faludi, 2005b), which should be subjected to further research. In particular, the positive impact of polycentricity on mitigating spatial disparities is questioned (Meijers and Sandberg, 2006). iv) The issue of fair access to SGI is closely tied to the processes of polarization and peripheralization. In this respect, small cities play the key role in providing basic services and jobs for their hinterland.

3.3 The issue of scale The concept of polycentricity, viewed through the cohesive interpretation, is primarily understood in terms of its function for achieving balanced spatial development characterized by lower levels of spatial disparities. Attention will be paid to functional and spatial dynamics of the urban system, especially from the economic point of view focused on the centrality of spatial units stemming from their job and service functions. Polycentricity will be analysed by means of morphological and functional aspects (the relationship of cities and rural areas), i.e. as a structural process. Another essential feature of polycentricity is its dependence on the geographical scale. In this respect, territorial cohesion (through the concept of polycentricity) will be operationalized at the regional scale; in all cases, the basic spatial unit shall be the administrative territory of the municipality in all cases. Choosing this spatial level is based on the following reasons: i) Being aware of the functional-spatial relations in the regional context will enable a critical view on polycentric development promoted at European level. In other words, the spatial perspective of regional urban systems (which are a part of the national hierarchically superordinate system) can clarify the

relationships between major metropolitan areas forming a centre at European level and peripheral locations in the non-central positions. ii) Regionalization into spatial units representing functioning systems of regular activities introduces an important element into the debates about the importance of territorial cohesion: the functional region. At this spatial level, it is possible to assess material, or better said, physical (in terms of personal mobility as the indicator of interaction within the network of municipalities) nature of polycentricity and level of inter-regional disparities. iii) Regional level is essential for assessing the accessibility of SGI. Citizens of non-central municipalities are assigned to use basic services (health services, social services or retail) in other communities where to it is necessary to commute. The position of a municipality in the urban system and its own ability to supply basic services greatly determines the overall accessibility of services. iv) More or less pragmatic reason is the significant number of territorial cohesion analyses at European level, which were made within the ESPON platform (e.g. ESPON, 2005a, 2005b, 2007, 2012c). National and especially regional analyses are represented to a much lesser extent and require intensive attention from the analytical research.

4 STUDY AREA AND METHODOLOGICAL OVERVIEW

As previously mentioned, the empirical part is composed of five papers (four published/accepted and one in a review process) in scientific journals, where the main author is always the author of the presented dissertation (see Appendix 4). Each article focuses on a particular research objective18. Since each article includes a description of the specific methods that were used, this chapter serves only to form ideas about research techniques applied in the empirical part, which will ensure better coherence between theory, research objectives/questions and the empirical research itself. The whole territory of the Czech Republic was chosen as the study area (Paper 2, Paper 3 and partly Paper 1), with a closer focus on the South Moravian Region (Paper 4, Paper 5 and partly Paper 1). The whole Czech Republic was chosen for the following reasons: i) taking into account the regional and hierarchical variability of functional dynamic of the territory, which would remain undiscovered within a single region; ii) sufficient area, in terms of size and population density, to analyse the relationship between polycentric development and intra-regional disparities; iii) a territory affected by both socialist central planning and the current territorial cohesion policy. In the choice of the South Moravian Region the following factors played a decisive role: i) the significant hierarchical structure of the regional urban system as a prerequisite for spatial polarization and ‘unwanted’ development; ii) simultaneous occurrence of strongly urbanized areas and sparsely populated areas - the importance of urban-rural relationship for occurrence of regional disparities; iii) urban system under the influence of the Brno metropolitan area - suitable for assessing the significance of ‘borrowed size’ and ‘agglomeration shadow’ concepts; iv) relatively good availability of secondary data. First, since features of territorial cohesion bear a striking resemblance to the socialist central planning, the similarity between the currently proclaimed territorial cohesion policy and traditions of directive central planning is described (Paper 1) on the example of the Czech Republic and the South Moravian Region. For mutual comparison of the two systems, primarily the key documents concerning spatial planning from both periods are used. While the documents ‘Principles and Standards of Physical Planning’ and the ‘Physical Plan of Brno Settlement Regional Agglomeration’ are used to discover analytical practice of socialist spatial planning highlighting spatially balanced distribution of urban centres through the so-called ‘central settlement system’, current territorial cohesion discourse permeates (within the spatial planning framework) the documents ‘Spatial Development Policy of the Czech Republic’ at the national level and ‘Spatial Development Principles’ at the regional level. Definitions of urban agglomerations and population centres in both periods serve for a comparison of spatial planning approaches which reflect policies with a similar understanding of territorial cohesion principle.

18 The dissertation adopts, almost unchanged, the form of the published articles. Minor modifications can be seen in the formal structure in order to maintain a unified graphical editing of the work. Individual figures and tables, as well as the footnotes, are numbered separately in each article. Due to a relatively strong overlap of quoted sources, only one list of references, which includes resources from both the introductory theoretical part and individual articles, is provided at the end of the study.

The next part of the research is aimed at assessing the relation between the normative concept of territorial cohesion (understood as an ideal representation of space - see Lefebvre, 1991), particularly the spatial configuration of ‘polycentric urban systems’ (Kloosterman and Musterd, 2001; Hall and Pain, 2006), and actual individual spatial practices of contemporary society. The methodology is based mainly on an analysis of secondary work commuting data. First (Paper 2), development of the number, importance and relative position of centres within the urban system of the Czech Republic, during the two decades (1991, 2011) is described. Using Multiple Linkage Analysis (MLA) (Holmes and Haggett, 1977; Van Nuffel, 2007), work commuting data are reduced to a quantity necessary for identification of centres. Time comparison provides essential information on urban system development and allows to consider changing spatial routines and practices and to assess the degree of polarization (poly/mono- centricity levels) in terms of spatial and economic attractiveness. The current structure of regional urban systems, their mutual comparison in terms of poly/mono-centric character, and the association between urban system configurations and intra-regional disparities are evaluated in the next stage (Paper 3), using the Czech Republic as a case study. Original identification of regions is inspired by the method delimitating Travel- to-work-areas (Coombes, 2002), which enables regionalization respecting inner functional coherence of regions stemming from a degree of relational integrity of territories. Again, commuting to work is used as the input indicator. The levels of morphological (based on the rank-size distribution using the number of jobs) and functional (the degree of reciprocity of work commuting flows) polycentricity are examined for each defined region. From the socioeconomic and demographic perspective, intra-regional disparities are measured by traditional variability measures (the Gini coefficient and standard deviation). Finally, disparities are associated with the configuration of urban systems. Then the evaluation of the SGI accessibility, in the context of spatial and functional linkages within a territory, is applied to the area of the South Moravian region. The role of small towns in the urban system is assessed in terms of their service function and with regard to increasing the polarization of Czech territory (found out in previous stages of the research). Besides data on provision of services, the significance of towns is determined by focusing on the position in the urban system, the tourist and economic potential and transport accessibility (Paper 4). Using multinomial logistic regression, the typology of small towns is interpreted in the context of ‘borrowed size’ and ‘agglomeration shadow’ processes that have been recently often discussed within the fields of urban and economic geography (Fujita et al., 1999; Burger et al., 2015; Meijers et al., 2015). The final part of the empirical research (Paper 5) uses the same data set (service function) but focuses on the overall SGI accessibility and its potential impact on demographic (population aging) and socioeconomic (attractiveness of the area in terms of migration and housing or characteristics of employment) structure of municipalities. The central question is whether and to what extent, the imbalanced SGI accessibility has a negative impact on territorial cohesion. Calculation of SGI accessibility is inspired by Maier et al. (2010), which dealt with modelling of spatial attractiveness. The impact of accessibility on the above-mentioned municipal characteristics is assessed through multiple linear regression. The issue of accessibility is discussed within the context of the territorial cohesion discourse and in

54 connection with Rawls's theory of justice and the epistemological concept of spatial justice (Soja, 2010).

5 EMPIRICAL ANALYSIS OF TERRITORIAL COHESION

Paper 1: European territorial cohesion policies: Parallels to socialist central planning?

JIŘÍ MALÝ, ONDŘEJ MULÍČEK

Moravian Geographical Reports (2016), 24(1)

ABSTRACT Contemporary EU territorial cohesion policy presents some striking reminders of features of socialist central planning. The objective of socio-spatial solidarity aimed at balanced spatial development is a core principle of both spatial planning doctrines. Reviewing key planning documents, this article compares territorial cohesion discourses in terms of their normative and analytical natures in order to critically evaluate the uniqueness and novelty of the current modern concept. In spite of ideological contradictions, a commonly-shared realization of the importance of urban agglomerations as specific integrated spatial units and the need to improve living conditions in disadvantaged areas, are crucial characteristics for both spatial planning policies. Moreover, analytical spatial planning procedures are based on similar methods and lead to nearly identical results concerning the spatial pattern for one specific case settlement system (the South Moravian Region, Czech Republic). In this respect, the currently- emphasized territorial cohesion discourse is familiar to that in former socialist areas in Central and Eastern Europe. Based on these findings, spatial planning authorities should learn from the past in reflecting on the limitations and advantages of spatial planning in the socialist era.

Keywords: territorial cohesion, socialist central planning, settlement systems, European Union, Czech Republic

1 Introduction ‘Territorial cohesion’ has become a conceptual buzzword often quoted in European regional and spatial planning policies. It is a frequent subject of theoretical discussions concerning balanced socio-economic development, as well as a goal of planning and decision-making practices. The broad thematic scope and an underdeveloped analytical apparatus, however, make this concept rather elusive in terms of its operationalization and evaluation. It seems there is no single definition of territorial cohesion; instead, it is used as an umbrella term covering several purpose-built conceptual frameworks and approaches.

The term ‘territorial cohesion’ appeared in official EU documents for the first time in 1997 in the Amsterdam Treaty, with regard to the importance of services of general economic interest (SGEI). Here, the declared access to SGEI is understood as the cornerstone of territorial cohesion, but without any detailed specifications (Sauter, 2008). Later, the concept has become part of the regular reports on economic, social and territorial cohesion. The Third Cohesion Report defines territorial cohesion in a rather normative manner as a state of balanced development, reducing existing disparities and territorial imbalances (EC, 2004a). A likely more meaningful statement, however, is that ‘people should not be disadvantaged by wherever they happen to live or work in the Union’ (EC, 2004a, p. 27). Corresponding with this formulation, Martin and Ross (in Davoudi, 2005) suggest that the territorial cohesion concept ‘spatializes’ some variety of so-called biographical risks, such as unemployment, disability, poverty, etc. In other words, an individual’s life chances reflect not only his or her position within the system of social interdependencies but, at the same time, their position within the structure of territorial interdependencies. As Molle (2007, p. 84) points out, territorial cohesion is ‘a situation whereby people and firms are not unduly handicapped by spatial differences in access to basic services, basic infrastructure and knowledge’. The concept of territorial cohesion enunciated here echoes significantly the ideas of spatial justice understood as the ‘fair and equitable distribution in space of socially valued resources and opportunities to use them’ (Soja, 2009, p. 2). Employing a critical spatial perspective, research attention is drawn not only to the qualities of particular places and territories, but, more implicitly, to their organization in physical, socio-economic and political space. As places of work and living do not exist as isolated geographic entities, place-based qualities and opportunities stem from the complex networks of territorial interdependencies mentioned above. Each particular urban system involves a specific arrangement of territorial interdependencies reflecting, among many other factors, political strategies articulated in the form of planning doctrines and policies. Spatial planning can be therefore regarded as an important ‘platform’ translating essentially political concepts of territorial cohesion into the worlds of everyday urban activities. Planning interventions usually follow normatively-defined narratives of a territorially coherent society, applying tools related to the spatial (re)distribution of valued resources. The general aim of this paper is to look more closely at the conceptualization of territorial cohesion in two distinct periods of modern history. The contemporary operationalization of the territorial cohesion concept implemented in EU countries will be compared with the central planning doctrines endorsed by socialist regimes. Such an historical excursion could shed light on the currently proclaimed territorial cohesion concept. Special attention will be paid to the political discourse framing the territorial cohesion concept in both of the periods, and particular spatial policies which shape the functional geographies of interdependent places and territories will be examined. The paper indicates that territorial cohesion is an evolving concept deeply rooted in its spatial and political contexts. The comparison then focuses on a Czech case study (the South Moravian Region), as the changing socio-economic and political conditions, as well as the development of relevant planning tools, will be critically explored with the aim of disclosing the scale of conceptual shifts and their imprints on planning practice.

2 The conceptualization and operationalization of territorial cohesion 2.1 Territorial cohesion in current EU planning doctrine Achieving territorial cohesion is currently one of the main objectives of EU regional policy. The territorial dimension was officially attached to the goals of economic and social cohesion by the Lisbon Treaty (EC, 2007b). Thus, today, regional policy represents economic, social and territorial cohesion policy (Cohesion Policy). As a shared competence between EU and its member states, territorial cohesion aims at a coordination of policies with spatial impact to ensure integrated territorial development (Faludi, 2013). From this normative and theoretical perspective, the Territorial Agenda of the EU 2020 defines territorial cohesion as ‘a set of principles for harmonious, balanced, efficient, sustainable territorial development. It enables equal opportunities for citizens and enterprises, wherever they are located, to make the most of their territorial potentials’ (EU Ministers responsible for Spatial Development, 2011, p. 2). Based on this quotation it can be assumed that territorial cohesion is characterized (besides its political nature) by a strong spatial planning dimension that includes an aspect of social and spatial solidarity. This general and ambiguous definition, however, results in a number of different interpretations and reflections about the sense and relevance of the territorial cohesion concept (see e.g. Davoudi, 2005; Schön, 2005; Doucet, 2006; Evers, 2008; Servillo, 2010). The most noticeable confusion associated with the concept is a simultaneous promotion of the principle of solidarity and also the competitiveness of European regions and Europe as a whole. In this regard, Waterhout (2007) identifies the storyline ‘Competitive Europe’, stressing the need for a competitive European territory, which stands in contrast to the traditionally understood meaning of territorial cohesion emerging in the storyline ‘Europe in Balance’1. Given the purpose of this paper, the meaning of territorial cohesion emphasizing balanced development will be used. With regard to decision-making processes, the objective is to make ‘both sectoral policies which have a spatial impact and regional policy more coherent’ (EC, 2004a, p. 27). Additionally, more effective coordination of EU policies, member states’ authorities, private actors, planners and regional or local authorities is required. The system of multi-level governance should be able to manage functioning of various territories and enhance territorial cohesion (Finka and Kluvánková, 2015). Respecting the principle of subsidiarity and the so- called bottom-up approach, vertical and horizontal coordination between decision-making bodies at different levels and sector-related policies is supposed to secure consistency and synergy within the process of achieving territorial cohesion (EU Ministers responsible for Spatial Development, 2011). Reaching territorial cohesion should be based on an adaptation of development opportunities to the specific characteristics of a particular region. Thus, the diversity of regions is not ignored and is even regarded as a development potential (EC, 2008). Despite the awareness of the unique position and inner structure of each territory (notwithstanding its

1 In addition to ‘Europe in Balance’ and ‘Competitive Europe’ storylines, Waterhout (2007) also recognises the storylines of ‘Coherent European Policy’ and ‘Green and Clean Europe’.

59 delimitation), common territorial priorities for the development of the EU have been established by the Territorial Agenda of the EU 2020. From a planning perspective, territorial priorities reflect challenges for territorial development that cover a wide range of fields of interest (from demographic and social challenges to environmental risks and climate change). Consequently, the list of territorial priorities is very complex as well. ‘Balanced spatial development’ is seen as a key element of territorial cohesion and is predominantly associated with the structure of urban systems. The promotion of ‘polycentric development’ is therefore crucial in terms of avoiding the economic, social and spatial polarization of human activities (however, supporters of a competitive European territory scenario see polycentric development as a bridging concept of cohesion and competitiveness), although such an assumption lacks empirical verification (Veneri and Burgalassi, 2012; Malý, 2016). One the one hand, the potential of metropolitan areas to generate economic and social prosperity is recognized, and the attractiveness of the largest agglomerations for living, working and investment seems to be unquestionable. One the other hand, territorial cohesion discourse accentuates the complicated position of spatially excluded territories and suggests that ‘rural, peripheral and sparsely populated territories may need to enhance their accessibility, foster entrepreneurship and build strong local capacities’ (EU Ministers responsible for Spatial Development, 2011, p. 7). Realizing the importance of ‘territorial cooperation’, territorial priorities include improvement of ‘spatial connections’ (i.e. transport networks, communication technologies and infrastructure, cross-border relations, etc.) and strengthening ‘local economies’. Achieving territorial cohesion should also respect ecological and natural values. In this regard, ‘protection of ecological systems’ is also territorial priority. Territorial priorities defined by the Territorial Agenda of the EU 2020 should not be viewed as isolated goals. By linking them to strategy Europe 2020, Böhme et al. (2011) identify five territorial keys that can be understood as crucial issues promoted by the territorial cohesion concept: accessibility, SGEI, territorial capacities/endowments/assets, city networking, and functional regions. Based on territorial priorities it can be argued that urban systems and their functioning play a key role in spatial development. Due to the extent of the EU in terms of land area, territorial cohesion is characterized by strong scale-dependency. In the context of urban systems, the role of cities/towns is partially determined by the geographical level at which they act as centres. Nevertheless, according to the concept of polycentricity, centrality stems from nodal positions within the urban network and connections to other localities, irrespective of scale level. Access to centres is thus an essential factor when trying to improve living conditions in disadvantaged areas and to achieve more territorial cohesion (e.g. efficient public transport connecting rural municipalities to local towns, highway networks ensuring relations between regional capitals, or accessibility of the largest metropolitan regions by air transport). Besides transportation accessibility to centres (provision of SGEI and jobs), focus is put on easy access to communication services (broadband, mobile telecommunication) and energy networks. The principles of territorial cohesion are not ground-breaking. The promotion of balanced spatial development in order to reduce territorial disparities and more evenly distribute economic activities is deeply rooted in European policies. Interest in regional planning at the European level had begun to emerge during the second half of the 1960s. The formation of

60 regional policy and the beginnings of any actual applied principles of cohesion policy, however, can be dated to the 1970s. The establishment of a common regional policy partly related to the implementation of the Common Market, which was unable to balance the differences between regions, and partly to reducing economic disparities before the planned single currency project (George, 1996). With growing inter-regional inequalities due to the expanding membership base of the EU (mainly the post-socialist countries of Central and Eastern Europe), European spatial development has been seen as increasingly important. In the late 1990s, the principles of territorial cohesion began to form during the process of preparing the European Spatial Development Perspective. The promoted model, however, did not represent an innovative strategy in the context of spatial planning. Rather it was inspired by ‘l’aménagement du territoire’, a French tradition of spatial planning (Faludi, 2004). ‘L’aménagement du territoire’ was developed as a strategic spatial framework designed for eventual intervention by the public administration and was based primarily on a regional economic approach to spatial development (Faludi, 2009b). Economic changes in France in the 1960s (partially caused by the loss of the French colonial markets and the lowering of trade barriers within the European Economic Community), the increasing economic domination of Paris (at the expense of most other French regions) and consequently growing differences between the major cities and especially rural regions, were the main factors in adopting l’aménagement du territoire (Burnham, 1999). France, at that time a centrally-governed state, started to apply the policy of balanced spatial development with regard to geographical and social conditions in particular regions. The concept of territorial cohesion (as well as l’aménagement du territoire) emphasizes the development of disadvantaged areas or territories lagging behind. By supporting equal access to SGEI and jobs, the territorial cohesion policy applies the European social model to spatial planning strategies. Access to SGEI and jobs should be ensured for all citizens irrespective of where they live. Location of residence, economic and social activities and relations between them, are in themselves preconditions for a certain level of territorial cohesion. Thus, the spatiality of everyday human lives is closely linked to general welfare and social status. Promoting territorial cohesion adds a spatial justice dimension to European spatial policy (Davoudi, 2005). It seems that thinking about space has been evolving from economic and technical perceptions of space as a container to a recognition of spatial and social causality, something that Soja (1980) called the socio-spatial dialectic. From the perspective of critical geographies, however, the current political and economic organization of European space is one of the factors of spatial injustice. In contrast to territorial cohesion discourse, the epistemological concept of spatial justice in itself represents one of the critiques of capitalist economies. But in fact, no matter how truly socially motivated the promotion of territorial cohesion is, the aspect of social solidarity has become an integral part of EU spatial policy.

2.2 Territorial cohesion in socialist planning doctrines up to the 1990s Socialism can be regarded as a general term for a specific socio-economic and political structure that orders many aspects of societal functioning. Single-party political systems, strong ideological anchoring, state ownership of the means of production (land included), rejection of market principles and a wide preference for collective interests – these are some of the

61 distinctive characteristics of socialist regimes (Musil, 2001; Nedovic-Budic, 2001), when compared with capitalist societies in the period after WW II. In spite of such unifying symptoms of socialist order, however, there were numerous types of socialist societies, a variety of socialist frameworks stemming from specific historical legacies of the pre-socialist periods and from different adaptations of ideological premises to local milieu (Hamilton, 1976). As Musil (2001) points out, the socialist transformation was implemented in countries differing in terms of economic and urban structures, political institutions and cultural models. The seemingly homogenous space of socialist countries has thus to be grouped into several categories, enabling a proper description of applied planning strategies. Firstly, we can recognize the category of Central East European socialist countries, including East Germany and the former Czechoslovakia, i.e. regions with relatively high levels of pre-socialist industrialization and urbanization, as well as Poland and Hungary, representing countries with a heritage of deeper regional disparities. The second distinct group covers the agrarian or semi- agrarian socialist states of south-eastern Europe, including Romania, Bulgaria and the former Yugoslavia. The Soviet Union and the non-European, predominantly developing socialist countries, can be further distinguished as a third or even a fourth category within the outlined classification (Dingsdale, 1999; Musil, 2001; Sokol, 2001). The research interest here will focus mainly on the category of Central Eastern Europe, but still respecting the strong influence of political and planning paradigms emerging from the Soviet space during the post-war period. The political systems under the socialist regimes were tightly coupled with the economic and social ones. The interconnectedness was visible in particular in socialist industrialization which played the important ideological role of a flagship project, introducing not only economic but, at the same time, also social modernization (Mareš, 1988). It was precisely this strong ideological dimension that made socialist industrialization so different from other types of industrialization processes (Szczepański and Furdyna, 1977). Socialist industrialization was controlled through a strongly hierarchical central command planning system. The national economic strategy defined the basic framework for developing more specific policies for various sectors, and long- and middle-term economic priorities were set up on these decision- making levels to reflect the needs of the national economy as a whole (Hoffmann, 1994; Nedovic-Budic, 2001). The regional policies were generally given a lower priority, at least in the first two decades of socialist industrialization, which was understood as a comprehensive universal tool diminishing existing regional disparities. Regional plans were formulated as rather auxiliary documents channelling the geographical distribution of nationally-defined planning targets (Enyedi, 1990). The top-level regional documents took the form of urbanization strategies, which detailed physical arrangements at the nation-wide scale. Their effective design and scope followed to some extent historical legacies and national settlement specificities in their respective countries, as well as the modifications of political regimes since 1940s. We can recognize several distinct phases of socialist industrialization, having different impacts in terms of territorial interdependencies and regional disparities (Szczepański and Furdyna, 1977). During the period immediately after WW II and further into 1950s, the major effort was to restore national economies (Malík, 1976). The onset of industrialization followed the Soviet heavy-industrialization model, which was not accompanied by specific urbanization

62 strategies (Enyedi, 1996; Musil, 2001). The discourse of territorial cohesion was embedded primarily at the national scale, echoing the proclaimed equity between industrialization and socio-economic modernization. There was an ideologically-supported aim to develop new socialist industry outside of the traditional capitalist industrial cores, and accordingly some investments were allocated to less developed, more agrarian regions. Nonetheless, the bulk of industrial production remained stabilized in the pre-socialist locations in order not to weaken overall national economic output, manifesting the contradiction between the de-concentrating appeal of ideological visions and the agglomerating nature of economically-driven politics (Musil, 2002). In the case of Czechoslovakia, substantial political attention was paid to diminish the long-standing economic gap between the Czech lands and the Slovak territories, as well as between the northern and southern parts of the Czech lands. The displacement of the original German populations resulted in the need to repopulate peripheral regions of the country (Illner and Andrle, 1994). These issues were viewed as ad hoc planning assignments and not set into any wider planning concept. Musil (2001) summarizes the planning discourse at the time as driven only by economic strategies, applying centralized distributive tools in rather extensive ways while ignoring regional feedbacks. From the early 1960s the discrepancies between industrialization strategies and regional policies became the subject of deeper planning interest, as they caused problems both in terms of economic development and in terms of social cohesion. The territorial distribution of new industrial premises, for example, often did not correspond to the potential of local/regional labour markets (Mareš, 1988), resulting in a lack of the required labour force, long-distance commuting to work and emerging demographic imbalances in some industrial centres. Especially in the case of Czechoslovakia and Hungary, the spatial concentration of industry outpaced the tempo of the concentration of population, and this developed indirect urbanization strongly and complicated the rational distribution of (non-industrial) resources (Musil and Link, 1976). Planning attention thus turned towards urbanization strategies, promising to set up an optimal equilibrium between economic effectiveness and social goals. The assignment for such goals can be cited from the Czechoslovak period analytical document: ‘The basic task of our settlement regulation is to work out how to distribute effectively housing and amenities development in the context of a too scattered settlement structure and how to, at the same time, reach the optimal settlement standards for all inhabitants within the national territory. The only solution is to establish a network of economic, social and cultural centres within the settlement system which will be well accessible on a daily basis, providing economic conditions for the concentration of population. We have to locate new housing and amenities development in these centres.’ (Palla et al., 1962, p. 22). The first generation of these urbanization concepts was developed in Czechoslovakia and Hungary during late 1960s and 1970s, theoretically based on Christaller’s Central place theory (Ryšavý et al., 1992). They transferred the cohesion discourse from the national to a regional level, and at the same time they refused the political concept of cities as spatial containers for industrial production. Instead, urbanism was put back into the game through taking broader non- productive and service urban functions into account (Enyedi, 1996; Szelenyi, 1996; Wu, 2003).

This approach of ‘decentralized concentration’ (Malík et al., 1968) established the basic territorial framework for the centrally-planned allocation of investments. The delimited network of centres was normative and, to certain extent, utopian in nature. But by the middle of the 1970s, spontaneous processes had started to change the normatively- given spatial pattern of centres in a significant way (Musil, 2001). Many centres were developing more slowly than intended. In contrast, the hinterlands of some regional centres rapidly strengthened their positions within the national settlement systems. The criticisms of the central-place settlement system came from economic standpoints, together with more realistic analyses of urban processes, and set the stage for the birth of conceptually new urbanization strategies. These concepts took into account the existence of spontaneous urbanization processes, as well as the economic and demographic importance of emerging city regions and metropolitan areas (Musil, 2002). City regions (urban agglomerations) represented qualitatively new spatial units within the planning doctrines of those times. They were complex territories integrated through economic, social and transportation linkages, requiring qualitatively new definitions of cohesive territorial arrangements. Accepting the ‘universal’ nature of urbanization processes, the socialist planning doctrines were weakened in their normative stance. The detailed physical planning approach was slightly re-oriented towards the employment of more integrated spatial planning tools. The delimitation of ‘preferred urbanization axes’ and ‘integrated urbanized areas’ (Terplan, 1981) contextualized the cohesion concept in a more relational way.

2.3 Comparative scheme Territorial cohesion discourses are primarily contextualized by political and economic systems and by their instances in certain periods of time. While EU spatial development strategies have evolved in democratic societies characterized by a free market environment and the rapid qualitative increase of communication technologies and overall individual mobility, the previous socialist planning approach was based on a totally different political regime, characterized by strong central governance and a limited role for local authorities in spatial planning processes. With respect to the territorial cohesion concept, however, the normative principles of contemporary European spatial planning policy and those of socialist planning doctrines exhibit similar features concerning the aims and priorities of cohesive spatial development. Although there were different underlying ‘raisons d'être’ for the spatialization of socio-economic political narratives, the idea of balanced spatial development represents the common aspect of both planning approaches: the excessive concentration of (economic) activities should be counterweighed by the development of lagging areas that are disadvantaged in terms of access to resources. The principle of socio-spatial solidarity is thus embedded in both planning doctrines as a way towards a more just or effectively a more balanced spatial arrangement. From a spatial planning perspective, the political goal of balanced spatial development is achieved via specific interventions into the functioning of a settlement system, attempting to counterbalance the uneven distribution of resources. The planning action is thus oriented mainly toward the support of the settlements centres outside of the economically most advanced areas. Generally, small and medium-sized towns are frequent objects of planning interventions in

64 order to create a stable network of centres which would ensure the efficient use of their strengths, through coordinated cooperation (EC, 2008). According to EU spatial policy, cooperation between regional and local centres by the sharing of functions and provisioning of services contributes to less territorial concentration and more balanced development (EC, 2007a). Similarly, socialist central planning emphasized the role of centres in which basic public amenities are concentrated and where residents of particular hinterlands can satisfy their claims and rights to education, health care, social care, etc. (Musil, 2001). These centres should be spatially distributed as evenly as possible. In spite of distinctive urban system theories which serve as a framework for settings of spatial planning strategies and the delimitation of centres (see the empirical part of this work, below), a focus on daily-based access to services and jobs characterizes both planning traditions. There is a strong de-concentration bias underlying the normative discourse of contemporary European spatial planning, as well as in the ideologically-framed socialist doctrines. Planned de-concentration, however, often conflicts with the more spontaneous concentrative nature of many social and economic processes. Although de-concentration of economic activities is one of the main aims of current EU territorial cohesion policy, the impact of metropolitan regions in terms of global competitiveness and their role in economic development is considered as crucial (EU Ministers responsible for Spatial Development, 2011). Highly-urbanized areas enjoy agglomeration economies, the advantages of clustering particular activities, easier access to higher education and health or social care facilities, etc. Consequently, ‘this is reflected in the high level of GDP per head, productivity, employment and research and innovation activity relative to the national average in capital cities and in most other densely populated conurbations’ (EC, 2008, p. 5). Under socialist central planning policy, metropolitan and suburban processes were limited due to the equalizing and regulatory approach to spatial development (Hampl, 2005). Even such a strongly restrictive planning strategy, however, was not able to hide the specific functioning of the largest urban areas. Reflecting the strength of regional agglomerations, socialist planners realized the imperfection of administrative spatial boundaries and the importance of complex territorial frameworks including broader spatial relations. Thus, the concentration of people and economic activities into growth poles (especially industrial agglomerations) gained its conceptual utilization, leading to an increasing focus on highly urbanized areas (Musil, 2001). A certain duality in the planning paradigms spanning between concentration and de-concentration benefits, can be thus pointed out as a feature inherent in both doctrines under study.

2.4 Case study methodology Having compared socialist and contemporary territorial cohesion discourses, we can argue that they share significant common features. The similarities can be found mainly in the spatio- political normative narratives framing the planning goals. What still remains unclear, however, is the extent to which these narratives are (and were) reflected in analytical practices of spatial planning. The current principles of EU spatial policy are translated into national spatial development strategies and planning tools. In the Czech Republic, the empirical focus of this paper, the form of spatial planning documents follows the hierarchy of particular administrative territorial units. The EU territorial cohesion priorities are taken into account in the Spatial

Development Policy of the Czech Republic: ‘… a planning tool that sets up requirements and frameworks for detailed specification of planning tasks’ (MMR, 2015a, p. 11). As a national document, the Spatial Development Policy concerns the issues of cohesion at a rather general level, particularly accenting the integrated development of cities and regions (reflecting spontaneous concentrative processes within metropolitan areas), as well as the polycentric organization of the settlement system (reflecting the normatively-defined goal of balanced spatial development). The general framework set by the Spatial Development Policy is developed into more concrete goals and measures by the Spatial Development Principles. This is the spatial planning document at the regional level and it must respect the Spatial Development Policy in order to ensure the vertically-binding interconnection of spatial planning documents. These documents (together with ad hoc studies of regional settlement structure) provide the information about analytical procedures that are based on the discourse on territorial cohesion. Correspondingly, socialist documents on spatial development serve as the source for understanding the practical application of socialist central planning principles. The structure of the historical planning documents under study is analogous, in many ways, to that of the contemporary materials. The documents produced by the state Research Institute of Construction and Architecture were examined to interpret knowledge of planning measures at the national scale. The document ‘Principles and Standards of Physical Planning’ (VÚVA, 1979) played the role of an historical counterpart to the contemporary Spatial Development Policy document. Analogically, the Physical Plan of the Brno Settlement Regional Agglomeration (Terplan, 1985) provided information concerning reflections of socialist national-wide policies in this specific regional context. Reviewing spatial planning documents, comparing analytical approaches and their impact on the spatial arrangements of territories, this empirical study reveals the parallels and dissimilarities of EU and socialist spatial planning. The starting point for the empirical analysis deals with the national-scale level, in an effort to compare patterns of normatively-defined territories, where intensive development is (was) expected to take place. The first step in the analysis is based on the planning policies coping with concentration processes. We argue, that the socialist map of ‘growth poles’ (urban regional agglomerations) is very similar to the contemporary normative delimitation of metropolitan regions. The (dis)similarity of policies intended to even out spatial imbalances is examined at the regional level in the second stage of this empirical analysis. This stage follows the normatively-planned de-concentration measures. Because the lower hierarchical level was crucial for the application of socialist de-concentration policies, the study region (namely the South Moravian Region - NUTS3) was established as the basic spatial unit for this part of the study. It was selected primarily due to the structure of its settlement system, including a variety of centres in terms of population size and economic importance, and hence it serves as a relevant model when describing urban hierarchies. The planned structures of the urban centres in the 1980s and the situation at present can now be compared.

Fig. 1. Geographical location of the South Moravian Region. Source: Czech Statistical Office (2014c); authors’ elaboration

The South Moravian Region is situated in the south-eastern part of the country and is characterized by high economic potential, especially given by the strong position of its regional capital Brno in the national economy (see Figure 1). Moreover, its strategic location stems from its proximity to the metropolitan regions of Prague, Vienna and Bratislava. Regarding the spatial relations and functioning of its settlement system, Brno plays a key role as the administrative, economic and cultural centre of the region (Mulíček and Toušek, 2004; Kunc et al., 2012). The importance of Brno (380,000 inhabitants) is further increased by the relatively small sizes of other centres (approximately 35,000 inhabitants of the second largest city Znojmo), and its central position with reference to spatial context and routing of transport infrastructure (Kraft et al., 2014).

3 Empirical analysis and findings 3.1 National level – urban regional agglomerations and the metropolitan areas Socialist urbanization strategies were characterized by a continuous evolution of the normative approaches to spatial development. The first theoretically-based conceptions were questioned and modified by approaches emphasizing spontaneous urban processes and the importance of highly-urbanized areas. With respect to analytical planning practices, the goal of decentralized concentration was initially expressed by the so-called ‘central settlement system’. The insufficient ability of the central settlement system to react to natural urban processes gave rise to strategies taking into account relatively spontaneous concentrative metropolitan processes.

In order to regulate these urbanization trends, urban regional agglomerations were delimited at the national planning level. They were conceptualized from the late 1970s as the elementary backbones of the national settlement system. The spatial delimitation of urban regional agglomerations is depicted, together with the metropolitan regions which were delimited as the target areas of integrated territorial investments (with respective to integrated development territorial plans), in 2014 (see Figure 2). Despite the long historical gap in development (these two distinct layers of metropolitan regions/agglomerations are almost 40 years distant from each other), there are just minor changes of overall geographic pattern. The number of delimited metropolitan regions is slightly higher in 2014 when compared with the 1976 proposal, as the Northern Bohemia urbanized belt was divided into two polycentric metropolitan regions and the Mladá Boleslav region emerged driven by the presence of a strong economic actor (Škoda Auto). Having accounted for the changes in spatial extent (which are sizable in the case of some metropolitan regions when compared to the socialist proposal), no other major structural variances which would distinguish the geographic logic of both delimitations are observed.

Fig. 2. Comparison of urban regional agglomerations (URA)a delimited in 1976 and ITI/IPRU metropolitan regions (MR)b delimited in 2014. a 1- Northern Bohemia URA, 2 – Plzeň URA, 3 – Prague URA, 4 – Liberec URA, 5 – České Budějovice URA, 6 – Jihlava URA, 7 - Hradec Králové/Pardubice URA, 8 – Brno URA, 9 – Gottwaldov (Zlín) URA, 10 – Olomouc URA, 11 – Ostrava URA b A – Karlovy Vary MR, B – Ústí nad Labem/Chomutov MR, C – Plzeň MR, D – Prague MR, E – Mladá Boleslav MR, F – Liberec MR, G – České Budějovice MR, H – Jihlava MR, I - Hradec Králové/Pardubice MR, J – Brno MR, K - Olomouc MR, L – Zlín MR, M – Ostrava MR Source: VÚVA (1979), MMR (2015b); authors’ elaboration

The question then is how much the similarity of spatial patterns stems from the affinity of socialist and contemporary planning discourses. The urban regional agglomerations were delimited during the 1970s as a kind of planning response to the gradual and rather spontaneous emergence of complicated inter-urban relations in the hinterlands of large Czech cities. These territories emerged from the traditional conceptualization of local daily-urban systems organized through flows-to-work in secondary sectors. The VÚVA period analytical documents (1979) point out the functional division of labour developing between particular towns and municipalities within agglomerations. In particular, the rise of employment in the tertiary sector in metropolitan cores formed a qualitatively new spatial configuration. The analytical and planning discourse thus had to shift from quantitative urbanization issues towards a more integrative approach able to grasp the functional diversity of urban regional agglomerations. The socialist integration discourse was different from contemporary concepts of integrative planning, however. It understood agglomerations as urban systems with an internal hierarchy of particular centres and municipalities. Different functions and development strategies were normatively assigned to them in order to reach a desirable development of the agglomeration as a whole. Although there were several proclaimed targets of planning measures (among them environmental, infrastructural and facilities issues), the coordination of economic and housing policies was of the highest priority. As the extent of sprawling suburbanization was restricted during socialism, the spatial balance between normatively allocated production and housing functions was one of the most important planning goals within urban regional agglomerations. The political and planning narratives at the base of the delimitation of present-day metropolitan regions differ in terms of scale. In contrast to the situation in the 1970s, there is a strong embeddedness of national planning actions in European political discourse. Re- territorialisation, as well as the re-scaling of regional policies and planning measures, have become emerging issues within this discourse (MMR, 2015b). Bearing in mind the socio- economic significance of European metropolitan regions, it is not surprising that urban/metropolitan dimensions receive the foreground of planning attention. Thirteen metropolitan regions were identified in the Czech Republic in 2014, ordered in two hierarchical levels – the metropolitan areas of Integrated Territorial Investments (ITI), and urban agglomerations of Integrated Plans of Territorial Development (IPRU). As mentioned above, the ‘top-down’ delimitation of socialist urban regional agglomerations is similar to the rather ‘bottom-up’ delimitation (there was no binding national methodology), managed at the regional scale. Disregarding the internal hierarchy, these metropolitan regions coincide with socialist urban regional agglomerations not only in terms of spatial delimitation but also in terms of political grounding. Just as in Czechoslovakian socialist planning discourse, they represent shifting scale, singularities emanating from existing spatial planning categories; they are proposed as the areas where integrated planning tools are to be applied, joining sector-based planning measures and financial schemes under a single strategic framework. The strategic documents giving reasons for the purpose of individual metropolitan regions employ concepts and issues highly comparable with socialist narratives. They respect the extraordinary (economic) position of urban cores, as well as the transforming nature of secondary

69 metropolitan centres. Contemporary metropolitan plans, however, do not attempt to prescribe fixed functions to the metropolitan centres, as the initial ethos of socialist planning was substituted by a more networked and participatory approach. The metropolitan areas were considered in both periods under study as growth poles, where special planning measures had to be employed in order to ensure economic performance and, at the same time, internal coherence.

3.2 Regional level – urban centres Besides realizing natural concentrative tendencies, socialist nation-wide spatial policies applied the approach of ‘decentralized concentration’, relevant especially at lower hierarchical levels. At the regional level, a strictly normative approach to delimitations of centres (the initial definition of the central settlement system from the 1960s) was relaxed in the 1980s. Although the places of concentration of human and economic resources were still seen as growth poles, their potential to generate economic profit was tightly connected to their specific spatial, economic and social advantages supporting concentration tendencies (Terplan, 1981). As a result, the revised concept of the socialist settlement system was based on the delimitation of a hierarchical settlement system reflecting the main functions of potential centres and the broader spatial context. In the 1980s, centres were defined by using information about a settlement’s functions and its regional significance (Terplan, 1985). The importance of centres was generally based on two indicators. The primary characteristic concerned the main types of residential, job and service functions. The second and rather additional indicator assessed the regional significance of centres by comparing the size of its respective micro-regions. A criterion of minimum functional size was adopted to determine settlement centres. In some relevant cases, a centre was represented by the organic integrity of more than one municipality. In other words, intensive mutual relations between settlement centres, expressed by mutual work commuting flows, served as a supplementary indicator to determine the final list of 338 centres (from 7,511 municipalities in 1970) at the national level, and 43 centres in the case of the South Moravian Region (in its present delimitation). Using the two previously-mentioned indicators, the defined centres were divided into four main categories (see Table 1). The first category (A) represents basic settlement centres characterized by a low frequency of units and a large inner differentiation of the significance of centres. Besides the capital Prague, which is the only macro-regional centre, this category includes meso-regional centres with distinctive levels of significance. Basic settlement centres are predominantly centres and other larger settlements of the highest-level administrative units – regions – in their former delimitation. Secondary settlement centres (B) are micro-regional centres with relatively great importance for their hinterlands. Supplementary settlement centres (C) are micro-regional centres typified by looser relations between functions of centres and their regional significance and by more variability in a centre’s development potential. Spatial context and other features concerning position within the settlement system are important for planning intentions. Other settlement centres (D) play the role of sub-regional centres with local significance.

Tab. 1. Categories of centres based on their function and regional significance (in 1985) and their presence in the South Moravian Region (SMR). Category Settlement centres Hierarchical level Sub-categories Number (SMR) Macro-regional A 0 A1 strong 1 A Basic Meso-regional A2 medium 0 A3 weak 0 B1 very strong 0 B Secondary Micro-regional B2 strong 1 B3 medium strong 2 C1 medium 2 C Supplementary Micro-regional C2 weak 4 C3 very weak 10 D Other Sub-regional D 23 Sum 43 Source: Terplan (1985); authors’ calculations

Tab. 2. Categories of centres based on production and service potential (in 2014) and their presence in the South Moravian Region (SMR). Settlement centres Sub- Number Category (1985 Category Positional typology * (hierarchical level) categories (SMR) delimitation) 1 Supra-regional - 1 a A1 2 I. 1 f B2 Regional 3 II. 6 f B3 4 I. 6 b, c, f C1 Sub-regional 5 II. 6 c, e, f C2 6 I. 9 a, c, d, e, f C3 Micro-regional 7 II. 8 c, d, f, g D 8 I. 12 b, c, d, f, g D Local 9 II. 5 b, c, d, f D Sum 54 * a = the core of the Brno metropolitan area (BMA); b = within BMA (strict delimitation); c = within BMA (looser delimitation); d = part of other agglomeration; e = networked with other centres; f = autonomous; g = periphery centre. Source: UAD Studio (2014); authors’ calculations

With respect to the principles of the current territorial cohesion policy, the basis for regional spatial planning policies in the Czech Republic is represented by the Spatial Development Principles (USB, 2015) and the Territorial Study of Settlement Structure (UAD Studio, 2014). In the case of the South Moravian Region, a municipality with production and service potential is understood as a centre (UAD Studio, 2014). Centrality is thus determined essentially by potential job opportunities, causing work in-commuting flows, and by the potential of service functions affecting in-commuting flows in terms of different types of services (retail, cultural facilities, social care, health care, administrative functions, etc.). Moreover, the importance of residential functions is also considered. Nevertheless, a purely quantitative approach including evaluation of the occurrence and prevalence of specific types of services or functions is not

71 applied. Besides quantitative indicators (static and dynamic), settlement context and the embeddedness of a specific centre in broader functional relationships are taken into account. Focusing on the delimitation made in 2014, the final number of centres in the South Moravian region is 54 (from 647 municipalities). The regional capital Brno is identified as a supra-regional centre (the only one in the region). Then, there are regional centres, sub-regional centres, micro-regional centres, and local centres (see Table 2). The hierarchical categorization is complemented by the positional typology of each centre, however. In this context, a centre could be the core of Brno metropolitan area (BMA), situated within BMA (strict and looser delimitation), part of another agglomeration, networked with other centres, autonomous, or a periphery centre.

Fig. 3. Delimitation of settlement centres in the South Moravian Region by UAD Studio in 2014 and Terplan in 1985. Source: UAD Studio (2014), Terplan (1985); authors’ elaboration

In general, methodological approaches to the delimitation of centres in both time periods show similar features. Primarily, both analyses are based on quantitative methodology concerning jobs and the services and residential functions of municipalities. Although this could result from the limited availability of municipal data, the focus on jobs and service functions is a traditional way to identify settlement centres. Secondly, the position and significance of centres within the settlement system is based on horizontal linkages and the potential integrity of particular territories. But, in fact, emphasizing the importance of relational aspects with regard to the identification of centres is a typical concern of current analytical approaches. Thirdly, although the comparison of results could be problematic due to different scales and methods employed

72 in both analyses, the distribution pattern of centres varies to a smaller extent and the main centres preserve their importance (see Figure 3). The categories of centres determined in 1985 have been assigned to the nine categories created in 2014. A regional analysis from 2014, however, defines a large number of categories and thus the comparison should not be overestimated. It serves especially as a graphical visualization and summary of the principal outcomes of the empirical investigation.

4 Discussion and conclusions This paper examined territorial cohesion discourses characteristic for spatial planning doctrines in two historically distinctive periods of time. Comparing the current territorial cohesion concept pervading EU Cohesion policy and the spatial planning strategies at lower geographical (administrative) levels with socialist planning doctrine in the Czech Republic, the work reveals remarkable similarities not only in the spatio-political normative narratives but also in analytical practices of spatial planning mechanisms. In spite of ideological contradictions between both spatial planning doctrines emphasizing social solidarity within spatial contexts, there is a shared principle of spatial development strategies. Uneven development is thus understood as a consequence in part of spatial inequalities resulting from various levels of territorial potential. Such a common ground is essential for the subsequent interpretation of particular narratives and analytical procedures. The EU territorial cohesion concept could be understood as a political goal and also as a tool designed to ensure ‘spatial’ solidarity across the EU territory. Despite several attempts to clarify the concept (e.g. Faludi, 2004; Servillo, 2010), its multidimensional character does not allow a simple definition. The abstract meaning of the concept becomes clearer when territorial cohesion is reflected in spatial planning strategies and the structure of urban systems is questioned. In that case, supporting small and medium- sized towns as local centres, as well as metropolitan regions as growth poles of the EU and national economies, is a typical practical application of the territorial cohesion concept. Natural concentration processes increase the importance of the largest agglomerations while peripheral localities experience dramatic losses of population and economic power. A territorial cohesion policy aims at overcoming territorial disparities by finding solutions to the adverse situations of disadvantaged regions (EC, 2008). Similarly, the goal of socialist central planning was to eliminate economic and social disparities between cities and rural areas (Malík, 1976). The issue of spontaneous concentration (or de- concentration) processes was purposely mitigated (and ignored) in the initial phase of the socialist period in Czechoslovakia. To a certain extent, however, the ‘failure’ of the central settlement system approach enabled the formation of a more complex and integrated approach to spatial development. The realization of the economic significance and specific functioning of urban regional agglomerations was a first step to modify the previous strict normative planning strategy into a more contextualized approach, respecting the distinctive qualitative nature of specific spatial units. Thus, both territorial cohesion discourses are characterized by a relatively substantial concentration/de-concentration duality. Reviewing the analytical procedures and outcomes of both spatial planning doctrines we can argue that spatial pattern of agglomerations/metropolitan areas in the Czech Republic has not changed in a significant way. Nevertheless, the socialist approach was based on a

73 slightly different understanding of urban agglomerations: although an agglomeration consisted of several spatial units, linkages between them were planned as vertical policies with the focus on coordination of economic development and housing. On the other hand, the current delimitation of ITI metropolitan regions respects increasing de-concentration processes, the functional specialization of secondary centres emerging in mutual horizontal linkages between the spatial units, and the growing importance of the core city or entire metropolitan region for more distant municipalities (i.e., the larger area of ITI metropolitan regions as a consequence). Regardless of distinct internal processes, metropolitan areas are seen as specific spaces (territories in EU rhetoric) with great impacts on national (European) development, especially in terms of economic prosperity, and as units exceeding traditional administrative boundaries and requiring integrated planning tools. Information about the analytical elaboration of the normatively-defined goal of supporting small and medium-sized towns is provided by the delimitation of centres at the regional level. The socialist delimitation worked with the absolute importance of centres (defined by jobs, service and residential functions), and the broader context including relations with surrounding municipalities played only a supporting role. Greater emphasis is put on the capability of centres to create their own catchment areas in the current delimitation. Due to the enormous stability of settlement systems, however, and despite transformation processes in the Czech Republic in the 1990s, the outcomes of both delimitations show a considerable degree of accordance. With regard to the type of centres in terms of their functions and desired (planned) development, any contradictory distinction between socialist and current spatial planning is mainly based on related economic systems and modes of production. In the case of centres located in peripheral and rural areas, socialist planners accentuated agricultural functions with a strictly defined hierarchy depending on specific localization related to agricultural land and potential consumption. The development of other centres was closely linked with industrialization and agricultural mass production (Malík, 1976). Today, the centrality of peripheral centres is related to a broader spectrum of functions and activities and is more dependent on the position of the centre within the urban network. In terms of spatial planning policies, the territorial cohesion concept does not represent a completely new spatial planning strategy, at least in the former socialist countries and especially in the Czech Republic. In spite of different ideological backgrounds, planned balanced spatial development is typical for both territorial cohesion discourses. As a common objective, spontaneous concentration processes should be counterbalanced by the growing prosperity of peripheral and rural regions. While socialist policies aimed at ensuring prosperity by direct investments in the production functions in central settlements, current EU regional policy intends to enhance the adverse situations of peripheral areas by strengthening local entrepreneurship, especially through the investments in the form of subsidies from the EU structural funds. The centralist top-down approach of socialist spatial planning has been replaced by a more decentralized system, characterized by a certain level of autonomous decisions concerning the spatial development of particular territories and a more limited power of the state apparatus. In contrast with the socialist regime, contemporary spatial planning policy is applied within a distinctive socio-economic context: a free market environment; intensive mobility; international trade; and globalization influences. In this matter, the EU goal

74 of territorial cohesion seems to play the role of a socially-motivated ‘rescuer’ of areas not profiting from the capitalist economic system. As natural concentration processes continue, however, with the increasing importance of the largest agglomerations and metropolitan regions projected in the support of growth poles (ITI), one can seriously doubt improved cohesion for the most disadvantaged areas. This leads us to essential questions concerning the functioning mechanisms of a capitalist economy tightly connected to the concentration of wealth, production or even ideas into a relatively small number of key development centres. Learning from the mistakes of socialist spatial planning associated mainly with the partial ignoring of regional and local specificities, current European spatial planning policies should be aware of the problems related to the strict following of normative concepts and grand narratives. Urbanization processes emerge in a rather natural (or at least politico-economic) way, and thus spatial planning practices should be based on complex and integrated planning concepts and instruments. Instead of a non-effective application of a normatively-defined spatial redistribution of centres, contemporary territorial cohesion discourse places an emphasis on the advantages resulting from spatial diversity and the particularities of unique localities. Nevertheless, EU territorial cohesion policies build on grand narratives, including access to SGEI, polycentricity, or territorial capital with the purpose of continuous economic prosperity. Territorial cohesion practice should not be limited only to a growth and competitiveness rhetoric, but rather the regional diversities stemming from the varieties of European territories should be brought to the forefront of interest. In the context of the negative historical experiences of the former socialist European countries with central planning mechanisms, scepticism towards top-down spatial planning equalizing policies is a legitimate concern. As a multi-scalar and multidimensional concept, territorial cohesion attempts to grasp all of the issues linked with regional development – without a real awareness of the complicated realization of this task with respect to the site-specific character of spatial inequalities. In this regard, understanding territorial cohesion as a place-based approach, even though it disregards to some extent the complexity of local development and requires different scenarios and practices in different spaces, seems to be a crucial interjection in order to move forward the effectiveness and comprehension of the territorial cohesion concept.

Paper 2: Moving towards more cohesive and polycentric spatial patterns? Evidence from the Czech Republic

JIŘÍ MALÝ, ONDŘEJ MULÍČEK

Manuscript (under review in European Urban and Regional Studies)

ABSTRACT Territorial cohesion discourse represents normative and ideological vision of ideal spatial practices that should be applied in order to ensure balanced and polycentric spatial development. However, spatial diversity of existing territorial settings often diverge from political representations of spaces. Using the Czech Republic as a case study, the paper focuses on the development of urban system hierarchy by analysing the changes of work-commuting flows in the post-socialist period. The results show that the configurations of everyday spatial routines and functioning of real urban systems differ from political imaginaries to a great extent and are moving towards more polarized pattern.

Keywords: territorial cohesion; polycentricity; significant commuting flows; urban systems; Czech Republic

1 Introduction The notion of balanced spatial development has been part of EU regional policy since its inception in 1970s when European Regional Development Fund was created. Increasing importance of overcoming territorial disparities has been gradually recognized due to the growth of economic and social inequalities stemming from EU enlargements (EC, 2008; Molle, 2007). Discussion about necessary enriching of EU policies by spatial aspect led to the adoption of European Spatial Development Perspective (ESDP) in 1999. ESDP is now viewed as a trigger of following efforts to give rise to genuine European spatial planning policy (Waterhout, 2008). Despite the lack of EU formal competences in the field of spatial planning and non- binding character of ESDP, open discussion about spatial planning topics at the EU level is certainly ambitious and ‘particularly new and stimulating task’ (Vanolo, 2010, p. 1303). In this fruitful period, territorial cohesion has become one of the most pursued concepts. And after a short time, territorial cohesion was introduced by the Treaty of Lisbon as a new goal of the EU development strategy (EC, 2007b). Although territorial cohesion is not defined in the Lisbon Treaty, its official political acceptation makes the concept most relevant. As a result, not only EU regional policy but sectoral policies as well should be strongly focused on territorial challenges and territorial dimension of overall development (EC, 2008; EU Ministers responsible for Spatial Development, 2011).

The territorial cohesion is undoubtedly politically driven concept which provides only limited reference points to its operationalization and analytical treatment. It is primarily the heterogeneous discourse of spatiality what makes the issues of territorial cohesion so resistant to simple and all-encompassing analytical grasp. Henri Lefebvre contends that ‘spatial practice, representations of space and representational spaces contribute in different ways to the production of space according to their qualities and attributes, according to the society or mode of production’ (Lefebvre, 1991, p. 46). Retaining this perspective of Lefebvre’s triad territorial cohesion might be conceptualized as a specific representation of space combining rather ideological imaginary (Church and Reid, 1999) of just and balanced spatial practice with more or less systemized planning knowledge. It embraces partial normative concepts like polycentric development or urban-rural partnership in which specific spatial configurations are highlighted as the ideal scenes for practicing everyday spatial routines. The spatiality of present European society, however, stems from complicated relations between variety of production and consumption spaces produced under different territorial settings and scales. The situatedness of everyday life (Cox and Mair, 1988) routinely embedded in the micro-regional scale sometimes interferes with political/planning concepts attached to the regional or national scale (Sýkora and Mulíček, 2009) deepening the gap between the tissue of individual spatial practices and the planning imaginaries. As discussed later in the paper concept of polycentric development represents a meaningful contribution to the corpus of territorial cohesion theory and, at the same time, spatial issue par excellence. Polycentricity discourse inevitably involves discussion on planning impacts to spatial structure of European urban hierarchies as well as topologies of inter-urban relations (Peters, 2003). The framing aim of the paper is to approach the issue of territorial cohesion from this very perspective of polycentric discourse in order to expose possible discrepancies between normative cohesion policy agenda and the spatial practices imprinted into the development and functioning of the real urban system. The work-commuting flows will be employed here as a proxy of spatial practices producing and shaping the relational space of multiple ties between particular geographic places. There is not an isotropic space under examination – it includes centres which centrality stems from their prominent position within the pattern of flows as well as non-centres dependent on the central places. The pattern of relational spatial practices is not fixed in a long-term horizon – relational space of everydayness evolves and transforms reflecting multiple spatialities of economic and societal processes. Employing the datasets which depict work-commuting in 1991, 2001 and 2011 enable following the transformation changing relation space of national labour market in the Czech Republic. The number, importance and relative position of centres develops as the spatial pattern coming from the late socialist era gets overlaid by the new scene of socio-economic spatial practices. Some research questions can be raised here addressing the dichotomy between planned and real space of cohesive policies. Do the ongoing changes make for more polycentric organization? To what extent spontaneously produced settings fits with normative imagination of polycentric and balanced space? What are the possible implications for cohesion-planning theory? The paper is organized as follows. The next chapter discusses theoretical grounds of territorial cohesion concept with special attention paid to the polycentric spatial development.

In order to outline the context of the empirical analyses the specifics of the Czech urban system development have been described further in the chapter. - A particular attention has been paid to the era of socialist central planning as well as to the subsequent period of economic and political transformation. It is followed by the section describing data and methods employed in the analyses. Analytical part of the paper addresses the development of the job centres pattern between 1991 and 2011 providing deeper view into the transforming hierarchies within the network of central places. Finally, the analytical results will be critically discussed in relation to contemporary European polycentric development discourse.

2 Theoretical and historical background 2.1 Territorial cohesion as a normative political concept The normative meaning of the territorial cohesion concept is still ‘fuzzy’ and rather ambiguous (Davoudi, 2005; Evers, 2008; Faludi, 2004). Davoudi (2005) argues that such quick adoption of relatively vague term is related to its projection of positive perspective. In this respect, Davoudi (2005, p. 433) states that ‘ambiguity becomes an advantage because people of different convictions can sign up to them without committing themselves to any particular interpretation or any particular application’. The result is a frequently used ‘buzzword’ in the context of ‘EU spatial development policy’ (Sarmiento-Mirwaldt, 2015; Schön, 2005). Faludi (2005b) suggests that territorial cohesion is primarily a political concept whose purpose is to build consensus. From a political perspective, clearly defined concept with precise meaning would not be helpful in reaching agreements and compromises. Despite his scepticism about concept’s definability, Faludi (2009b, p. 24) summarizes that ‘territorial cohesion refers to a situation whereby policies to reduce disparities, enhance competitiveness and promote sustainability acquire added value by forming coherent packages, taking account of where they take effect, the specific opportunities and constraints there, now and in the future’. One of the most striking features of territorial cohesion is the dichotomy between ‘competitiveness’ and ‘cohesion’ (Vanolo, 2010; Zonneveld, 2000). These two interpretations, defined by Waterhout (2008) as storylines of ‘Europe in balance’ (traditional cohesion objectives characterized by redistribution of financial resources that aims at reducing regional disparities and supporting development of peripheral and disadvantaged regions) and ‘Competitive Europe’ (EU global and regional competitiveness, capital invested into the metropolitan cores and optimization of European territorial structure), originate from diversity of spatial planning traditions across EU member states. Consequently, heterogeneous mix of approaches to regional development through spatial planning concepts is reflected in the complicated definition of territorial cohesion as an umbrella concept for the whole EU territory. In order to interconnect conflicting storylines of ‘cohesion’ and ‘competitiveness’ the spatial planning concept of polycentric development was introduced to EU regional policy primarily by ESDP (EC, 1999).

2.2 Polycentric development: crucial part of the territorial cohesion discourse Polycentric development has been strongly supported since the adoption of ESDP at the end of the past century. As it is stated in ESDP (EC, 1999, p. 20), ‘the concept of polycentric development has to be pursued, to ensure regionally balanced development […] pursuit of this

79 concept will help to avoid further excessive economic and demographic concentration in the core area of the EU’. Thus, polycentric development stands for balanced and sustainable development, i.e. the crucial goal of territorial cohesion. Additionally, polycentric development has been also related to the EU global competitiveness as a form of clustering of people and economic activity outside the existing core areas. As a result, more balanced spatial development would lead to more competitive Europe and its regions (Waterhout, 2002). From the policy-making perspective, polycentric development has gained its popularity due to stressing more balanced EU territory and also the importance of economic performance for territories lagging behind. Besides the complexity of polycentricity emerging from cohesion-competitiveness dichotomy, polycentric development has multiple interpretations when the issue of scale is questioned. Territorial Agenda 2020 (EU Ministers responsible for Spatial Development, 2011, p. 7) aims ‘at polycentric development at the macro-regional, cross-border and also on national and regional level in relevant cases’. In this context, realizing scale-dependency of the polycentricity concept is an essential factor for its better understanding (Taylor et al., 2008; Vasanen, 2013). With regard to relational space of everydayness which forms the core of regional urban settings, polycentric development is characterized by mutual interactions, cooperation and complementarity between similar sized cities (ESPON, 2014b). Polycentricity is simply based on daily social and economic interactions that occur in a territory without a clear dominant centre (Kloosterman and Musterd, 2001). It is a spatial arena of daily human lives including commuting to work, school or services. Thus, integrated network of similar sized cities with intense mutual linkages is a basic assumption for creating polycentric urban region (Davoudi, 2003; Kloosterman and Musterd, 2001; Parr, 2004). From this viewpoint, the role of small and medium sized towns is significantly dependent on spatial position within the particular urban system. While being a part of metropolitan area might lower the importance of smaller towns (agglomeration shadow discussion – see e.g. Burger et al., 2015), localization of such towns in rural and less densely populated areas increases their centrality because of providing essential services for the surrounding municipalities (Van Leeuwen and Rietveld, 2011). Green Paper on Territorial Cohesion states (EC, 2008, p. 6) that ‘the role these towns play in providing access to services including the infrastructure necessary to invest in the adaptability of people and enterprises, is key to avoiding rural depopulation and ensuring these areas remain attractive places to live’. However, growing importance of metropolitan areas attracts population and economic activities from more distant areas and contributes to ongoing process of depopulation of peripheries (in some regions). Local towns situated in peripheral and rural areas are therefore losing their central position for its surroundings. In comparison to monocentric urban systems, polycentric urban regions ensure plurality of choices regarding job opportunities and service functions (Malý, 2016). While the monocentric structure suffers from excessive concentration of human activities that negatively affects environment, causes congestions and infrastructure degradation, polycentric systems should benefit from economies of scale and scope and clustering of economic activities (Kloosterman and Lambregts, 2001; Meijers, 2008). In other words, polycentric development is an opposite of polarisation resulting from disproportionate concentration of economic

80 activities into one or a little number of centres (Ezcurra et al., 2005; Maza and Villaverde, 2004). Polarisation could emerge when the role played by local centres (mostly in peripheral areas) in economic and social development is decreasing and dominance of regional capital is growing. In presented study, polycentric development is therefore understood as a process characterized by growing density of urban centres’ network with enhanced role of medium- sized and small centres, while polarisation represents the development leading to excessive concentration of economic performance into a limited number of centres and deterioration of the others.

2.3 The Czech urban system: Introducing the spatial context for the empirical analysis The empirical part of the research is put into the spatial and historical context of the Czech urban system. Czech Republic (as a part of the former Czechoslovakia) belonged during the second half of the 20th century to the group of Central East European socialist countries that were marked by relatively high levels of pre-socialist industrialization and urbanization, at least when compared to the rest of the socialist block (Dingsdale, 1999; Malý and Mulíček, 2016). The socialist industrialization, introduced in the near post-war period, was meant as a prominent political and economic goal that was, however, accompanied by significant socio-spatial impacts (Malík, 1976). Industrialization policies implied not only the support of pre-socialist industrial urban centres; they aimed, at the same time, to strengthen the economic (industrial) potential of less developed regions and towns mostly situated in non-industrialized and rural parts of the country. However, Mareš (1988) points to the fact that the potential of many local labour markets was not often sufficient to ensure proper functioning of industrial production sites. Centrally command socialist economy therefore had to solve symptoms of a lack of compliance between industrialization and urbanization policies – e.g. poor availability of qualified workforce, long commuting distances, or gender and age imbalance in the industrial towns (Musil and Link, 1976). As a consequence, socialist spatial planning practice started to pay more attention also to the non-productive aspects of urbanization and the specific spatial policy known as ‘central settlement system’ was set up. Theoretically based on Christaller’s Central place theory the planning scheme reflected non-productive and service urban functions that are of great importance for optimal housing standards and social cohesion (Enyedi, 1996). The role of normatively defined centres across the national territory was to provide basic public amenities where inhabitants from particular hinterlands can meet their needs for education, health care, or social care (Musil, 2001). At the end of a socialist period, the Czech urban system was largely comprised of a dense network of hierarchically structured urban centres which played a key role of local centres for rather well-defined hinterlands, especially in the peripheral and rural regions. The economic and political transformation period beginning after 1989 has been characterized by demise of central command economy and by emergence of rather spontaneous and intense spatial processes within regional and local labour markets under the free-economy conditions. Tertiarization of the Czech economy accompanied by massive deindustrialization, especially in mining, heavy industry and manufacturing industries (Osman et al., 2015), redrew considerably the map of urban system dynamism. Decline in employment or even closing down

81 of many factories negatively affected the socio-economic situation of mostly small and medium-sized towns (predominantly situated in peripheral border areas – see Malíková et al., 2015) whose development was often closely related to fully operational state-owned companies (Ouředníček et al., 2011). On the other hand, progressive tertiary sector and foreign investments activities were concentrated into the inner parts of the largest cities or suburban locations of urban agglomerations (Domalewski and Baxa, 2015; Sýkora and Bouzarovski, 2012). The paper aims at analysing Czech urban system development by focusing on relational space of work-commuting flows and interpreting the results in the light of contemporary territorial cohesion discourse that, in its normative sense, shows remarkable similarities with socialist spatial planning doctrines in the former socialist countries in Central and Eastern Europe (Malý and Mulíček, 2016) despite its ideological and governance contradictions (Dabrowski, 2014). From this perspective, the Czech urban system in the period 1991 – 2011 represents an analytical platform where the impacts of European neoliberal policies and principles of the free market can be well documented. While the relational space described by 1991 dataset still largely refers to the socialist configuration of the urban system, the 2011 situation reflects 20 years of multifaceted post-socialist development, advent of new spatial processes including suburbanization and metropolization (Ouředníček et al., 2015) as well as emergence of neoliberal planning discourse.

3 Methodology 3.1 Study area: the Czech urban system The Czech urban system is dominated by Prague, the economic and administrative centre of national importance with the population more than 1,200,000 inhabitants; the second-tier large cities of Brno and Ostrava show significantly lower positions in terms of population and economic performance (Figure 1). The spatial pattern of daily activities is, however, shaped by high number of medium-sized and small towns which play a crucial role in the Czech urban system morphological as well as functional organization. In this context, the Czech urban system shows a higher incidence of minor centres that represent pivotal spatial units in terms of wider urban system functioning (Sýkora and Mulíček, 2009). The national urban system can be thus subdivided into a large number of functionally integrated and relatively self-contained micro-regional units representing elementary territories of daily commuting to work, school and services, which number range from about 100-160 (Klapka et al., 2014; Halás et al., 2015) to 250 (Sýkora and Mulíček, 2009) depending on the particular method of delimitation.

Fig. 1. Urban system of the Czech Republic – municipalities with 10,000 and more inhabitants. Source: Czech Statistical Office (1991, 2001, 2011); authors’ processing

3.2 Data and methods The empirical part of the paper employs census data on work-commuting flows1 between municipalities in years 1991, 2001 and 2011. Municipalities in the Czech Republic are the smallest administrative units and basic territorial self-governing communities of citizens that are governed by elected representatives. There were 5,768 municipalities in the Czech Republic in the year 1991, 6251 in the year 2011 (Czech Statistical Office, 1991 and 2011) - the increase of the number of municipalities after 1991 has been primarily caused by the efforts of some municipalities to become independent (predominantly forcibly merged municipalities in the socialist era). Accordingly, the urban system is rather fragmented at the LAU 2 level (Sýkora and Mulíček, 2009) and the matrix of inter-municipal work-commuting flows is very complex.

1 Inter-censual comparability of work-commuting data is affected by missing identification of workplace by one- third of the economically active employed persons in 2011. This could limit consequent analysis, especially when calculating with the absolute size of work-commuting flows between municipalities. Thus, the database of work- commuting flows in 2011 was adjusted in order to enhance its operability. Data adjustment involves the redistribution of persons with unidentified municipality of workplace. Based on the knowledge of their number for each municipality of their permanent residence, they were proportionally assigned to economically active employed persons in a particular municipality with known workplace. The chosen procedure obviously cannot reconstruct precisely the actual spatial distribution of jobs, but approximates the various work-commuting flows to their real size at least. Some degree of inaccuracy might be observed when focusing on individual work- commuting flows. However, general trends and relations within urban system hierarchy, which are the matter of concern for presented paper, are affected minimally.

Following the empirical purpose of the paper the work-commuting flows are to be employed to indicate centres within urban system. The analysis is thus focused only on those flows that represent meaningful bond between the pair of municipalities and the complete commuting matrix is purposively cut down to the reduced matrix of significant flows. As figured out in Table 2 the significant flows comprise approximately a half of total commuting volume in the Czech urban system. Significant flow might be defined here as the outgoing flow which proportion on the total volume of commuting from the municipality is significantly higher than the proportions of other, insignificant outgoing flows. Depending on the number of outgoing significant flows the municipality can be meaningfully assigned to one or more other municipalities which play the role of employment centre(s) in this respect. A municipality is regarded as a centre if it is a destination of at least one significant flow. The method for determining the number of significant flows comes from an application of multiple linkage analysis (Haggett et al., 1977; Holmes and Haggett, 1977) presented by Van Nuffel (2007) and Van Nuffel et al. (2009). Technique of significant flows determination is based on a comparison between ideal and real composition of the set of flows outgoing from a given municipality. The five highest outgoing flows were selected for each municipality and their percentage shares of the sum were calculated. Real distribution of flows was correlated with the five ideal configurations, each clearly indicating the number of significant flows (see Table 1). Finding the highest correlation between real and ideal configuration the number of significant outgoing flows was determined.

Tab. 1. Ideal configurations of outgoing flows.

Number of significant The share (%) of flow in the sum flows 1st flow 2nd flow 3rd flow 4th flow 5th flow 1 100 0 0 0 0 2 50 50 0 0 0 3 33 33 33 0 0 4 25 25 25 25 0 5 20 20 20 20 20

The analytical assets of this approach could be called into question for neglecting non- significant flows. The application potential, however, strongly depends on the analytical and interpretative framework of the research. The presented paper does not deal primarily with functional hierarchy of Czech urban centres where omitting the non-significant flows would be problematic. The primary focus is on the degree of polarisation within the national urban system and its development. That is exactly the research context in which (1) the significant flows approach enables to reduce effectively the relational complexity of the urban system as a whole and which also (2) co-defines the terms ‘centrality’ or ‘central place’ (rather binary terms referring to the fact that the municipality is a destination of at least one significant flow or not). The hierarchical scaling of delimited centres is not absent as it combines external indicators (job-size of the centre) with the internal ones (the number and volume of incoming significant flows). At the same time, it is necessary to take account also of the interpretational limits related

84 to the analytical method employed. The work-commuting flows represent just only a part of the totality of the relational space - there are more spatial practices like shopping or leisure time trips which co-produce the space of daily routines, whether monocentric or polycentric in its nature. Furthermore, the hierarchization of centres is rather hybrid in its nature, its interpretation is relevant just for given territorial and time framework. At first, basic characteristics of urban system using the number of significant flows as fundamental variable are presented (Czech Statistical Office, 1991, 2001 and 2011). Next, number of centres, significant flows and the total number of commuters are categorized according to the size (categories) of centres in terms of the number of jobs in order to assess the transformation of hierarchy of central places between the years 1991 and 2011. The number of incoming significant flows represents the hierarchical position of centres by reflecting a scope of the commuting zone. More precisely, the number of significant flows for each centre actually provide an information about the number of municipalities integrated to a centre by a strong functional relation involving journeys to work. The number of significant flows coming to the particular centre may be therefore equated with the number of municipalities assigned to this centre, i.e. the number of integrated municipalities. Comparison of two time periods has to cope with different number of municipalities in two stages (1991 and 2011) of the Czech urban system development. There was an intensive increase of the number of municipalities within twenty years of post-socialist development (from 5,768 in 1991 to 6,251 in 2011) mostly caused by disintegration of many large municipalities in which two or more previously autonomous settlements were joined together during socialist period. The municipalities established after 1991 are not included in the analysis. Keeping them in a calculation for 2011 would distort the comparability of the number of significant flows. The commuting flows matrix of 2011 was therefore adjusted to the spatial structure of 1991. The municipalities which ceased to exist administratively after 1991 (subsumed into other municipalities) are also excluded from the analysis. The change of the total number of commuters in significant flows is relativized by the number of economically active persons in the Czech Republic in both years in order to respect different economic population structure (smaller number of economically active persons in 2011 stemming from socio-economic and demographic changes).

4 Empirical analysis Determination of significant flows allows to identify centres; following the change of the number of centres, general findings on transformation of spatial organization of a particular urban system can be formulated. In the case of the Czech Republic, the number of centres dramatically decreased between 1991 and 2011 (Table 2). The number of centres in 2011 reaches about 60% of 1991 value. In other words, nearly half of centres in 1991 lost their centrality function. On the other hand, the number of significant flows and the total number of commuters within the urban system remain more or less stable. The national urban system thus gets more polarised as there is decreasing number of destinations of the significant flows while their total volume and number do not change. This is well documented by the growing average number of significant flows assigned to a centre (from 6 in 1991 to more than 10 in 2011) and average number of commuters commuting to a centre (from 627 to 1,031). Looking at values

85 of the year 2001, it can be argued that the change is quite continuous. Such a development strikingly points out the polarisation tendencies arising from spatial concentration of economic activities and jobs respectively. Using the EU spatial planning terminology, detected development of significant flows pattern might be considered the opposite of balanced and polycentric development. The question is, which municipalities enhanced their role as a centre within an urban system and which municipalities were not able to maintain their economic centrality.

Tab. 2. Work commuting significant flows in 1991, 2001 and 2011 in the Czech Republic. Centres and significant flows (SF) 1991b 2001a, c 2011a 2011/1991 number of centres 1,505 1,103 857 0.6 ↓ number of SF 9,323 9,030 8,968 1.0 ↔ total number of commuters (only SF) 943,003 864,630 883,212 1.0d ↔ share of SF of total volume of work commuting 0.55 0.50 0.52 1.0 ↔ avg. number of SF originated in a municipality 1.6 1.6 1.6 1.0 ↔ avg. number of SF directed to a centre 6.2 8.2 10.5 1.7 ↑ avg. number of commuters commuting to a centre 627 784 1,031 1.8d ↑ avg. number of commuters within one SF 101 96 98 1.0d ↔ a the calculation excludes municipalities established after 1991 in order to ensure comparison of time periods; b the calculation excludes municipalities annexed by other municipalities after 1991 in order to ensure comparison of time periods; c the calculation excludes municipalities annexed by other municipalities after 2001 in order to ensure comparison of time periods; d the change is relativized by the total number of economically active persons in the Czech Republic in each year, indicators are divided by the total number of economically active persons in a particular year and then the change is calculated. Source: Czech Statistical Office (1991, 2001, 2011); authors’ calculation

Categorization of centres according to their economic size in terms of the number of jobs provides necessary information for deeper understanding of the change of central places’ organization (Table 3). Although a decrease in the number of centres in particular categories may be affected by adapting of the labour market to lower number of economically active persons in 2011, relativizing of the number of centres in each category by the total number of economically active population yields the same results. Hence, a mutual comparison truly refers to changes in the spatial dynamics and hierarchies of socio-economic relations within an urban system. The most significant decrease of the number of centres can be observed in the category of the smallest centres (decrease by 50% between 1991 and 2011). This rapid decline of the economic importance of a large number of the smallest centres is accompanied by substantial fall of the number of significant flows and the total number of commuters coming to these centres.

Tab. 3. Centres of significant flows in the Czech Republic, the change between 1991 and 2011. Number of incoming significant flows Number of centres Total number of commuters (within SF) Size category of (number of integrated municipalities) centres (number of c jobs) 2011/1991 2011/1991 1991b 2011a 2011/1991 1991b 2011a 1991b 2011a abs. per 1 centre abs. per 1 centre 100,000< 4 4 1.0 ↔ 861 1,615 1.9 ↑ 1.9 ↑ 187,656 312,810 1.8 ↑ 1.8 ↑ 30,001-100,000 18 10 0.6 ↓ 971 1,033 1.1 ↑ 1.9 ↑ 166,444 143,583 0.9 ↓ 1.7 ↑ 15,001-30,000 33 26 0.8 ↓ 1,304 1,168 0.9 ↓ 1.1 ↑ 171,981 131,273 0.8 ↓ 1.0 ↔ 7,501-15,000 55 49 0.9 ↓ 1,470 1,579 1.1 ↑ 1.2 ↑ 147,315 137,064 1.0 ↔ 1.1 ↑ 3,001-7,500 145 97 0.7 ↓ 1,839 1,705 0.9 ↓ 1.4 ↑ 149,458 102,426 0.7 ↓ 1.1 ↑ 3,000> 1,250 671 0.5 ↓ 2,878 1,868 0.6 ↓ 1.2 ↑ 120,149 56,056 0.5 ↓ 0.9 ↓ All centres 1,505 857 0.6 ↓ 9,323 8,968 1.0 ↔ 1.7 ↑ 943,003 883,212 1.0 ↔ 1.8 ↑ a the calculation excludes municipalities established after 1991 in order to ensure comparison of time periods; b the calculation excludes municipalities annexed by other municipalities after 1991 in order to ensure comparison of time periods; c the change is relativized by the total number of economically active persons in the Czech Republic in each year, indicators are divided by the total number of economically active persons in a particular year and then the change is calculated. Source: Czech Statistical Office (1991 and 2011); authors’ calculation

As the overall change of the number of commuters and the number of significant flows is very small within the studied period, significant flows redirected from small centres have had to shift the whole structure of the centres’ hierarchy. Arguably, the largest centres with population size over 100,000 inhabitants increased their importance for a larger number of municipalities. This is evident when looking at the change of the number of work-commuting significant flows in case of the largest cities. These centres almost doubled the number of incoming significant flows (the number of integrated municipalities) and the total number of commuters during the two decades while the number of those centres remained stable (Prague, Brno, Ostrava, Plzeň). The share of significant flows assigned to the largest cities substantially increased. The aggregate number of significant flows for categories of centres providing 3,000 to 100,000 jobs is without a substantial change, however, the number of centres reduced in all these size categories. Thus, the number of integrated municipalities per one centre increased (a particularly steep increase occurred especially in the category of the second largest centres). Similar results can be seen for the change of the total number of commuters per one centre. Unlike other job-size categories of centres, the smallest centres are characterized by decreasing numbers of commuters per one centre – contrasting with the enlargement of the number of municipalities integrated to the small centres. Although general tendencies of the change of central places’ structure can be identified, more detailed perspective is required in order to better understand the spatiality of complicated commuting patterns and interactions.

Tab. 4. Work commuting characteristics according to the size categories of centres, the case of the Czech Republic. The share (%) of centres in particular typological categories (change of the Size category of number of incoming significant flows a, b/change of the total number of municipalities (number commuters within SF a, b, c) of jobs in 2011) being strong moderate stagnation moderate strong centres in 1991 and/or growth growth decline decline 2011 (>1.50) (1.11-1.50) (0.91-1.10) (0.51-0.90) (<0.51) 100,000< 50/25 50/50 -/25 - - 30,001-100,000 60/60 40/20 -/20 - - 15,001-30,000 -/12 42/15 46/19 12/50 -/4 7,501-15,000 6/4 31/20 43/25 20/49 -/2 3,001-7,500 6/10 24/10 31/12 26/39 13/29 3,000> 13/12 2/1 11/2 12/8 62/77 All centres 13/12 5/3 13/4 14/13 55/69 a the calculation excludes municipalities established after 1991 in order to ensure comparison of time periods; b the calculation excludes municipalities annexed by other municipalities after 1991 in order to ensure comparison of time periods; c the change is relativized by the total number of economically active persons in the Czech Republic in each year, indicators are divided by the total number of economically active persons in a particular year and then the change is calculated. Source: Czech Statistical Office (1991 and 2011); authors’ calculation

The change of the number of significant flows as well as the change of the total number of commuters can be categorized according to a degree of growth or decline (Table 4) in order to effectively complement the previous analysis. While most of the larger centres are characterized by strong (or moderate) growth of flows and commuters, with decreasing size of centres the

88 regressive trends seem to be prevailing. The centres within two largest size categories show solely increasing number of integrated municipalities. The next three size categories (3,000 to 30,000 jobs) represent mostly centres characterized by stagnation of the number of integrated municipalities and by the loss of the economic attractiveness regarding the number of in- commuters. More than one half of the smallest centres experienced strong decline in terms of the number of integrated municipalities as well as the number of in-commuters. Concentration of significant flows goes hand in hand with an increasing number of commuters travelling to work into the largest centres. This development takes place at the expense of medium-sized towns at the micro-regional level and especially local towns that are characterized by an extreme centrality function decline.

Fig. 2. Change of the number of integrated municipalities between 1991 and 2011, the case of centres of significant flows in the Czech Republic. Note: The category 'strong decline' includes 779 centres that lost all incoming significant flows after 1991. Source: Czech Statistical Office (1991, 2011), authors’ processing

A close association between the number of integrated municipalities (significant flows) and job size of centres (in 2011) is supported by a value of Pearson’s correlation coefficient r=0.94. However, some exceptions can be identified when a centre attracts the above-average number of significant flows compared to its job size (relatively ‘peripheral’ centres with wide hinterland attracting commuters from a large number of small municipalities) or when a centre is a destination of a lower number of significant flows compared to those expected based on its size (the influence of competing centres in the close proximity). The latter case is particularly visible

89 within the largest agglomerations, principally Prague and Brno metropolitan areas (Figure 2). While the core cities are characterized by strong growth of incoming significant flows, their hinterlands are comprised of declining centres that may suffer from the agglomeration shadow effect, i.e. reducing urban functions due to centripetal forces of higher-ranked metropolitan centre (see Burger et al., 2015; Partridge et al., 2009). From the late 1990s, the urban system development is strongly affected by intensive residential suburbanization processes within which the hinterlands of large Czech cities became the zones of large population increases (Sýkora and Mulíček, 2012). As the suburban immigrants largely keeps their jobs in the core cities, this development does not contribute so far to the strengthening of the economic attractiveness of minor urban centres in the hinterlands; instead the volumes of the work-commuting flows from the hinterland to the metropolitan cores have increased. The Western-type suburbanization processes were dampened during socialist period in many European socialist countries (Tammaru, 2001); post-socialist suburban boom at the end of the 20th century can be partly perceived as a reflexive correction towards development trajectories typical for the Western urban system. A number of authors agree on the fact that suburbanization represents major qualitative shift in the post-socialist Czech urban system development (Ouředníček, 2003; Čermák et al., 2009; Sýkora and Mulíček, 2012; Ouředníček et al., 2015). There is, however, a lack of consensus on the quantitative dimensions of suburbanization regarding the migration volumes as well as impacts to the commuting patterns. The ambiguity of suburban zones delimitation, methodological uncertainty related to analytical definition of suburbanization and only limited availability of data describing the socio-economic profiles of suburban migrants are the main reasons that make very difficult to build up a generalizing quantitative description of suburbanization in the Czech Republic. As for the impacts to the work-commuting relations, the analyses of Čermák et al. (2009) or Sýkora and Mulíček (2014) show that the residential suburbanization is followed (at least at the major Czech metropolitan regions) by relatively strong non-residential suburbanization of jobs, often located outside the minor urban centres in the hinterlands. The suburban migrants, largely well- educated and employed in tertiary and quaternary economic sectors, however, occupy only small part of those new suburban job opportunities. Imbalance between socio-economic profile of suburban migrants and the structure of the suburban jobs (Sýkora and Ouředníček, 2007) results in complicated patterns of metropolitan work-commuting flows, including cross- commuting and reverse commuting. In order to underpin the argumentation in the paper, we can conclude that (1) suburbanization unquestionably increase/redirect the volume/direction of significant flows from the suburbanized municipalities to the metropolitan centres; however, the quantification of these impact is out of the scope of presented analysis, (2) high spatial variability of the intensity of suburbanization processes both at the national (differences between particular metropolitan regions) and micro-regional (differences between particular municipalities within metropolitan region) level has to be taken into account as it can bias any generalization, (3) the post-socialist job-shrinkage in the small industrial centres in the hinterlands seems to be much more important factor reshaping the work-commuting flows pattern compared to suburbanization process.

Polarization within the Czech urban system is even more evident when the change of the total number of commuters is taken into account (Figure 3). While the number of significant flows remains stable for most of medium-sized centres, the volume of these flows (the total number of commuters) weakens. Such a trend is markedly occurred within the area of the Ostrava agglomeration, north-western part of the country and in the wider Prague’s and Brno’s hinterland. Regarding the notion of polycentric development at the national level, the domination of Prague is not equally counterbalanced by the growth of other largest agglomerations and regional centres. Moreover, micro-regional centres are losing their importance and the concentration of economic activities become more significant.

Fig. 3. Change of the total number of in-commuters between 1991 and 2011, the case of centres of significant flows in the Czech Republic. Note: The category 'strong decline' includes 779 centres that lost all incoming significant flows after 1991. Source: Czech Statistical Office (1991 and 2011), authors’ processing

5 Discussion and conclusion The updated version of Spatial Development Policy of the Czech Republic approved in 2015 introduces polycentric development as one of the eighteen national priorities concerning sustainable, integrated and cohesive planning. The document obliges national and regional planning authorities to ‘Support polycentric development of settlement structure. Create conditions for strengthening partnerships among cities and countryside and improve thus their competitiveness’ (MMR, 2015a, p. 16). These tasks are to be fulfilled by promotion of job opportunities and inter-municipal cooperation. The mentioned priorities are attuned with

European normative planning agenda as described in the introduction of the paper. The empirical analyses presented in the text, however, illustrate that normatively defined goals of cohesion (or polycentric policies respectively) diverge from the actual path of the Czech urban system development. The analyses carried out derive the spatiality of urban system from the patterns of work-commuting significant flows. In spite of the fact that labour-systems organization represents just a part of functional totality of national urban system, this approach allowed relatively simple but still comprehensive view of changes of relational practices shaping the space(s) of everydayness. Covering the period from 1991 to 2011 the analytical results indicate rather gradual re-centralization within relational space defined by commuting flows. The notion of re-centralization refers not only to more polarised structure of work- commuting centres but also to much broader shifts in the scale and spatial embeddedness of daily routine as the central places play a constitutive role in production and representation of space. Functioning of urban system as analytically depicted for the year 1991 was based vitally on the presence of numerous small job-centres. They can be viewed as materialized and rather persistent symptoms of specific spatiality having its roots in political and socio-economic milieu of socialist central command economy. Since the 1960’s normative urbanization strategies were developed in the former Czechoslovakia in order to harmonize spatial pattern of industrial job opportunities with housing and amenities development (Malý and Mulíček, 2016). The hierarchized network of job centres and their respective daily urban systems emerged; it was institutionalized and stabilized in time through planning narratives accenting the issues of jobs, homes and service of general interest accessibility. The dense pattern of socialist (industrial) job centres, many of them of rather small size, represented polycentric organisational model (when applying current terminology). The territorial cohesion stemmed from proximity which could be seen, at the same time, as a result of limited spatial mobility and planning rigidity. The matrix inherited from the times of socialism began to change in the mid-nineties. The empirical analysis shows the fall of the number of centres in 2001 and 2011 respectively while the total volume of commuting within significant flows remained virtually flat. The actual spatial pattern gets more polarized and perhaps less polycentric in comparison with 1991. The increase of importance of the decreased number of ‘successful’ centres is then a reflection of a competitive economic environment suddenly emerged during the socio-economic transformation occurred after 1989 in the post-socialist countries. There were, however, more impetus for change than shifting economic conditions resulting into the recentralization of investments, production and consumption at the expense of minor urban centres. It is necessary to mention deep social-spatial and socio-cultural changes accompanying the political and economic transition. Demand for owner-occupied housing and the sharp increase of the individual mobility can be seen as the triggers of massive suburbanization processes, which reshaped and rescaled the space of everyday relations especially within large metropolitan areas. In the case of large Czech metropolitan areas, a substantial numbers of inhabitants moved from the core city, however, without building any deeper work or consumption bonds to the secondary urban centres in the hinterland.

The development of urban system hierarchy is being far from normatively defined balanced spatial development emphasized by the territorial cohesion discourse. It seems that there is, at least in the Czech Republic, a lack of effective planning tools capable to operationalize the ‘grand narratives’ of current European planning debates. While the planning doctrines within socialist central command economy were tightly coupled with resources and investments re-distribution, the current bond between planning imaginations and grounded territorial policies appears much weaker. On the other hand, moving beyond the results of quantitative analysis several other issues can be addressed. The question is, to what extent the polycentric deficit deduced from the decline of the number of job centres is (or can be) counterbalanced by the increasing division of labour within the urban system. To what extent the more polarised pattern of centres relates to increased levels of personal spatial mobility and actual spatio-temporal habits of economically active population.

Paper 3: Impact of polycentric urban systems on intra-regional disparities: A micro-regional approach

JIŘÍ MALÝ

European Planning Studies (2016), 24(1)

ABSTRACT From the normative perspective of EU regional policy, polycentric spatial development has been seen as an effective tool to overcome regional disparities since the adoption of the European Spatial Development Perspective. However, the impact of a polycentric spatial structure on equal spatial development is ambiguous and still waits for its clarification. This paper examines the relationship by focusing on the local scale, where medium-sized towns play a crucial role as local centres in everyday human lives. Using functional regions of the Czech Republic, this paper defines polycentric urban systems at the micro-regional scale. Subsequently, the level of mono/polycentricity is compared with intra- regional disparities. Based on the results of the analysis, it can be argued that the connection of polycentricity with balanced spatial development is strongly determined by selected indicators. However, the findings do not provide evidence about the strictly positive impact of polycentricity on intra-regional disparities. In a light of these results, the conclusion is rather sceptical about the unique role of polycentricity in spatial development and due to the lack of similar studies, further research is recommended.

Keywords: spatial development; polycentricity; intra-regional disparities; Czech Republic

1 Introduction After the adoption of the European Spatial Development Perspective by EU member states at the end of the past millennium, polycentric development has begun to be seen as the core concept of EU regional policy (CEC, 1999). Reaching polycentricity has been even more emphasized by territorial cohesion as a goal of EU cohesion policy (Faludi, 2010). Polycentric development is now viewed as a process that contributes to lowering regional disparities and to the overall economic growth of the European territory (CEC, 1999; EU Ministers responsible for Spatial Development, 2011). According to the Territorial Agenda of the EU 2020 (EU Ministers responsible for Spatial Development, 2011), European regions are facing several challenges which should be transformed into potentials for sustainable and harmonious territorial development. The first territorial priority in this matter is to promote polycentric and balanced territorial development. It states that ‘where the most developed cities and regions within Europe cooperate as parts of a polycentric pattern they add value and act as centres

95 contributing to the development of their wider regions’ (EU Ministers responsible for Spatial Development, 2011, p. 7). Similarly in ESPON programme studies, polycentricity across cities and regions is seen as a necessary condition for a better distributed economic and social growth (ESPON, 2014b). Polycentric development is a vision not only of European territorial development, but also of many European countries (Meijers et al., 2005). European documents relating to spatial planning mostly address polycentricity in the context of large territorial units (macro-regions and cross-border regions), where the emphasis is put on the territorial competitiveness of the EU territory, improving the performance of cities in European and global competition, and promoting economic prosperity of second-tier metropolitan areas and cities that lag behind the largest metropolises - typically a situation in new EU member states where less-developed and large cities lie in the shadow of supported capitals (CEC, 1999; EU Ministers responsible for Spatial Development, 2011; ESPON, 2014). When the national scale and smaller regions are considered, the EU documents stress the importance of partnership between local centres and rural areas, and mutual complementarity of functions used on a daily basis which led to effective cooperation (ESPON, 2005a; EU Ministers responsible for Spatial Development, 2011). Thus, the principles of polycentric system functioning differ depending on a geographical scale. In the case of micro-regions, the absence of a stronger centre may result not only in a more polycentric pattern but also in particular dependence on an external centre. The popularity of polycentricity has grown in spite of (or perhaps because of?1) the unclear meaning of the concept (Molle, 2007). Polycentricity commonly refers to even distribution of human activities across space and the existence of multiple centres in one area (Kloosterman and Musterd, 2001). The core elements of polycentricity are urban systems since the polycentric structure of human activities may be best recognized through the spatial organization and functional interconnectedness of urban centres (Kloosterman and Musterd, 2001). It is a state of the urban system in which several nodes are interdependent and interact with one another via various types of flows (Parr, 2004; Limtanakool et al., 2009). In other words, the urban system lacks one dominant node and forms a network of cities. Polycentric urban systems may benefit from positive externalities associated with large agglomerations and simultaneously avoid negative aspects of urban sprawl (Parr, 2004). Thus, development of urban networks is the assumption for stimulating regional economic growth (CEC, 1999; ESPON, 2014b). A unique position of polycentric development within European spatial development is demonstrated by a plethora of studies and papers focused on the concept that can be divided into two types. The first are theoretical papers discussing the sense and importance of polycentric development (e.g. Kloosterman and Musterd, 2001; Davoudi, 2003; Parr, 2004; Faludi, 2005b) and the second are papers stressing the measurement and assessment of polycentricity mostly using specific case studies (e.g. Green, 2007; Meijers, 2007; Limtanakool et al., 2009; Goei et al., 2010; Burger and Meijers, 2012). Much less attention is paid to the actual connection of polycentric urban systems to regional disparities (see Meijers and

1 The ambiguity of spatial planning concepts sometimes becomes an advantage. Davoudi (2005) connects such positive political perception of polycentricity to the projection of positive perspective. She claims that ‘people of different convictions can sign up to them (i.e. ambiguous concepts) without committing themselves to any particular interpretation or any particular application’ (Davoudi, 2005, p. 433).

Sandberg, 2006; Veneri and Burgalassi, 2012). However, the merit of such studies is high since the feature of polycentricity to overcome regional disparities and contribute to a balanced spatial development is crucial. The present study is focused on functionally self-contained regions, that is, the micro- regional level, where intensive inter-urban interactions and everyday human activities occur. Regional and local centres are accelerators of development at this spatial scale (EU Ministers responsible for Spatial Development, 2011). The main aim of this paper is to assess the relation between urban system structure (the level of polycentricity) and the level of intra-regional disparities. Can we confirm the assumption that the higher the level of polycentricity is, the lower the internal disparities are? Is it possible to observe some changes depending on the population size of the region? Firstly, the paper focuses on urban systems theories which are strongly relevant in the introduction of the polycentricity concept. After a brief description of polycentricity, the crucial aspects of the relationship between polycentricity and regional disparities are subjected to a closer inspection based on the case study of Czech micro-regions.

2 Theoretical Background 2.1 From Central Place Model to Polycentric Urban Systems Cities have been seen as engines of economic growth since the first theories of organization of urban systems were formed (Christaller, 1933; Lösch, 1944). Cities have not been perceived as isolated units, but as centres providing specific services to surrounding localities. In theories of central place systems, the interactions between centres and rural areas were limited to trade and consumer behaviour (Berry and Pred, 1965). The principal centre’s market area contained surrounding municipalities which tended to be places of labour supply depending on the centre in terms of provision of goods and services. Moreover, attention was mostly paid to the size and distribution of cities. The central place model pointed to the hierarchical urban order with a clearly dominant centre and other lower ranked municipalities (e.g. Christaller’s Central place theory). At the intra-urban level, the monocentric city was represented by the well-known Burgess’s Concentric zone model (Burgess, 1925). While the urban core was the loci of economic activities, residential function was fulfilled by outer zones. Hierarchical organization of the city expressed by commuting flows from suburbs to the city centre resulted in nodal character of the model. At both geographical levels, the monocentric system is therefore characterized by core/periphery duality. Traditional central place models assumed that all urban systems have a rather monocentric character and the emphasis was put mainly on size-ranking and spatial distribution of cities. According to Parr (1987), the aspect of commuting is absent in the standard model of the central place system but it can be viewed as a result of an inequality in the demand and supply for labour at each hierarchical level. While the centre is characterized by an excess labour demand, the small places are characterized by an excess labour supply (Burger and Meijers, 2012). Based on the general urban systems theory, claiming that any urban system is formed by interdependent cities and functional interactions between them (Berry, 1964; Pred, 1977), central place models have been extended to general hierarchical urban models, which in some cases provide less strict definition of relations between the centre and hinterland (Berry

97 and Parr, 1988; Coffey et al., 1998 provide an overview). In other words, modified hierarchical models aim at a more realistic representation of spatial organization of urban systems. But the interdependence of cities does not have to represent only the centralized urban system. The interactions can be more balanced with less apparent hierarchy. Thus, the central place models are becoming unable to describe the changing urban systems at all spatial scales (Davoudi, 2003). New spatial and functional organization of urban systems becomes less hierarchically structured. Deconcentration of housing and economic functions within individual cities has substantially modified the traditional monocentric city structure. At the intra-urban scale, suburbs emerged into local centres which are now parts of the metropolitan area and fulfil not only the resident but also the economic function. Multidirectional commuting and the existence of multiple centres within the city gave rise to polycentric organization of the inner urban structure. However, the emergence of such networks is not limited to the intra-urban scale, but it also concerns inter-urban linkages (Meijers, 2007). With increasing geographical scope of social and economic activities (e.g. commuting and business relations), the functional linkages between original and new urban centres generate a more complex pattern of spatial organization of society (van der Laan, 1998; Kloosterman and Musterd, 2001). A variety of factors lies behind such evolution of urban systems (see e.g. Brotchie et al., 1995; van der Laan, 1998; Hall and Pain, 2006). The spatial organization of society has been transformed in particular by the structural process of globalization, which ‘involves the creation of complex world-wide webs in which economic actors such as producers, consumers, workers and investors are linked together in both physical and virtual ways’ (Kloosterman and Lambregts, 2001, p. 717). In the context of the post-industrial economy and globalization processes, the role of metropolitan areas and major agglomerations as economic, social and administrative centres within urban networks is growing. The significance of the urban centre is not expressed only by its size (e.g. economic, social and cultural), but primarily by its position within urban networks. The space of places still exists (people live, work or shop in specific places), but it is included in the broader and more dominant system of space of flows (Castells, 1996). These are the flows of capital, information or technology, or the interaction of different firms and organizations. From the perspective of relational space, the actors (people, cities or firms) do not exist per se, but their existence and functioning are conditioned and formed by interaction with other actors.

2.2 Conceptualizing Polycentricity Until recently, a clearly defined hypothesis explaining the functioning of urban systems (and replacing central place theory) had not been accepted (Camagni, 1993). Nevertheless, references to ‘non-monocentric’ forms of urban systems can be traced back to the first half of the twentieth century. In the context of city-region, Geddes (1915) used the word ‘conurbation’ for spreading of cities into the countryside. The concept of ‘dispersed-city’ was discussed few years later (Burton, 1963), and then the ‘network model’ was invented as a new model of spatial organization (Camagni, 1993; Capello, 2000). Although these concepts are based on different theoretical backgrounds, they have one thing in common. All of them were closely connected to the issue of what now is called ‘polycentricity’, the situation in the urban system

98 characterized by several similar-sized centres (or sub-centres, depending on the spatial scale) in close proximity that are connected by balanced and multi-directed flows of information, goods and people. Although there have been several attempts to clarify the concept of polycentricity (e.g. Hague and Kirk, 2003; Turok and Bailey, 2004; Lambregts, 2009; Sýkora et al., 2009; Groth et al., 2011), a widely acceptable definition is still missing. An important feature of polycentricity, which contributes to confusion regarding the concept understanding, is its dependence on the scale and different spatial arenas (see e.g. Hall, 2002; Taylor et al., 2008; Vasanen, 2013). The internal structure of cities is examined at the intra-urban scale. Attention is here paid to the processes of population and economic decentralization that have resulted in the creation of residential and commercial suburbs and new forms of spatial organization of society within the administrative city borders (Anas et al., 1998). At the inter-urban scale, the focus is on grouping of distinct cities without a clear dominant centre and with mutual interactions. The polycentric urban system at the inter-urban (regional) scale is defined by the concept of ‘polycentric urban region’ (Davoudi, 2003; Parr, 2004; Green, 2007). Polycentricity at the national and European levels is proclaimed by policy-makers as a path to combat regional disparities and to achieve more balanced development (CEC, 1999; EU ministers responsible for Spatial Development, 2011). At the national level, importance is attached to the development of medium-sized cities, which should contribute to reducing polarization between major cities, metropolitan areas and regional centres. Supporting polycentricity at the European level is mainly focused on metropolitan areas situated outside the current economic core2 that can restart their economic growth by forming networks and linkages with other cities (EU Ministers Responsible for Spatial Development, 2011). The aim is to create more globally competitive and economically strong regions, which would ensure the future of the European economy (Davoudi, 2003). It is evident that the issue of scale significantly affects the meaning and conceptualization of polycentricity. One of the possible results of such ambiguity is that a polycentric system at one scale may be considered as monocentric at another scale (Hall and Pain, 2006). The presented study addresses the micro-regional approach (reasons are discussed in the following section) and hence polycentricity will henceforth be conceptualized at the inter-urban scale. A polycentric urban system can be characterized by a greater number of centres located in the same area, none of which is significantly dominant. A polycentric urban system is distinguished by the distribution of economic and economically relevant functions over the spatial system in a balanced way, resulting in a less urban hierarchy (Meijers et al., 2005). However, the existence of multiple centres situated in close proximity does not mean that these centres are strongly functionally linked to each other (Albrechts, 2001). Thus, the polycentric urban system is expressed not only by a morphologically balanced structure, but also by infrastructural and functional complementarity of centres. The relational dimension of polycentricity is then characterized by balanced functional interaction between urban centres. Balanced relations could be simply imagined as functional linkages that are not directed at one centre, but that rather create mutual (reciprocal) and criss-cross interactions (Goei et al., 2010).

2 According to the ESDP, the European central economic area (so-called Pentagon) is defined by the metropolises of London, Paris, Milan, Munich and Hamburg (CEC, 1999).

Green (2007) presents a noteworthy study of functional polycentricity using an approach based on a formal network analysis technique. Networks comprise actors (e.g. cities, people, businesses and telephones) and relations between them (e.g. commuting roads and telephone lines), while these linkages are functional in nature. Green (2007, p. 2101) describes functional polycentricity by the statement that ‘a collection of nodes, be they cities, small businesses or people, must be functionally connected and balanced if they are to be considered a system’. Besides the balance of functional relations, Green (2007) adds the dimension of network density3 which reflects the level of connection between urban centres (or functional interdependency of urban centres). The level of urban system polycentricity can be the result of the morphological structure, functional linkages and network density. As Burger and Meijers (2012) have shown, although the level of morphological and functional polycentricity is interdependent to a certain extent, the level of network density is associated with different aspects of urban systems. When evaluating polycentricity, it is essential to distinguish different aspects of spatial organization of urban systems and consider them separately. As mentioned above, the highly integrated network of separate cities can be considered as a ‘polycentric (or polynuclear) urban region’ (PUR), a concept which has become widely discussed in the academic and policy debates (Kloosterman and Lambregts, 2001; Davoudi, 2003; Parr, 2004; Turok and Bailey, 2004; Green, 2007). Parr (2004) distinguishes a PUR from the other regions mostly by the small distance between cities, by no domination of any city size and by the intensity of the linkages and specialization of cities that is higher than that in other regions (e.g. Randstad Holland is considered to be the archetypal example of a PUR - see Lambregts, 2009). Defining ‘small distance’ and ‘intensity of linkages’ depends on the specific geographical context and population density. Also, the minimum size of a PUR in terms of market size and necessary human capital is questioned (Capello, 2000). Leaving aside methodological issues, the economic significance of PURs is seen in their specific spatial and functional structure. Meijers (2008) argues that PURs are able to develop new competitive advantages due to enjoying economies of scale, scope and complexity and avoiding agglomeration diseconomies, high land prices, congestions and environmental pollution. From a more social perspective, the specialization of centres should contribute to plurality of choices primarily in terms of job opportunities. Such assumptions about polycentricity advantages go hand in hand with a rather unjustified conviction of the European and some national policymakers about polycentricity and its positive impact on regional disparities.

2.3. Polycentricity as a Path to Overcome (Intra-)Regional Disparities? Up till now, research on polycentricity has been focused on conceptual and analytical issues. It can be argued that in the context of theory and methods, the knowledge about polycentricity has reached a certain level of saturation. This idea is also supported by Burger and Meijers (2012), who hope that polycentricity research is ‘no longer dominated by conceptual issues, but that focuses on whether the alleged benefits of polycentricity and polycentric development hold true or not’ (Burger and Meijers, 2012, p. 1145). Taking into account that polycentric development is a key part of EU cohesion policy (EU Ministers responsible for Spatial

3 Green (2007, p. 2084) has conceptualized network density as ‘the ratio of actual connections to total potential connections’.

100

Development, 2011), assessment of its real contribution to reducing regional disparities seems to be the most important part of the polycentricity research agenda. Paradoxically, the assumed vital aspect of polycentricity to overcome regional disparities is evaluated lastly. Although regional and spatial policies see the polycentric structure of urban systems as a situation contributing to lowering regional disparities, they do that without sufficient empirical basis. Meijers and Sandberg (2006) offer an empirical analysis of the relation between polycentricity and regional disparities in European countries. Their study shows that between most measures of polycentricity and regional disparities, there is no significant relationship. However, significant correlations in the opposite direction can be observed between some measures. In such cases, the monocentric urban system is characterized by less regional disparities. Their results are therefore in contrast to what have been expected. Not surprisingly the authors conclude that their ‘results call for a critical reflection on the assumed link between polycentric development and its contribution to diminishing regional disparities’ (Meijers and Sandberg, 2006, p. 18). Häzners and Jirgena (2013) used a similar approach to test the hypothesis about the positive impact of polycentricity on regional disparities. Their results show that higher polycentricity cannot be associated with less regional disparities. Another attempt to tackle this issue is the one made by Veneri and Burgalassi (2012). They investigate the relationship between the level of regional polycentricity and distribution of the key economic variables of performance (competitiveness, social cohesion and environmental sustainability) across Italian NUTS 2 regions. Especially social cohesion decreases with increasing level of polycentricity. In measuring income distribution, the authors find out that correlation exists between polycentricity and a more unequal income distribution (i.e. the opposite of expectations). Even the levels of productivity and environmental sustainability are not significantly correlated with polycentricity. In the end, the authors argue that a polycentric spatial structure can be hardly considered as an effective tool to ensure an equal spatial development concept, at least in the context of Italian NUTS 2 regions (Veneri and Burgalassi, 2012). The existing studies are rather sceptical about the crucial role of polycentricity in reducing territorial disparities, at least if a national and regional approach is considered. Certain limitations of the above-mentioned studies can be the comparison of simply administrative territorial units. NUTS 2 regions are divisions of countries used mainly for statistical purposes and their delimitation very often ignores functional aspects which form the spatial organization of socio-economic activities. This is obvious especially when the principles of functional polycentricity are taken into account. Thus, possibly strong cross-border relations may rapidly change the structure of NUTS 2 urban systems and consequently the level of polycentricity. In this respect, inappropriate territorial units disrupt also the morphological picture of an urban system structure. Even if the morphological approach is used, it is necessary to assess the level of polycentricity in areas that represent a functionally integrated system of urban centres and other municipalities. In the case of NUTS 2 regions, one or more centres belonging to one region can be in fact functional parts of a neighbouring region which can again change polycentricity measures. Notwithstanding the issue of delimitation of territorial units, these studies are important steps in building an empirical evidence base which is needed for polycentricity evaluation.

101

Unfortunately, additional research on linkages between polycentric development and territorial disparities is lacking. Especially the impact of polycentricity on intra-regional disparities has not yet been subjected to closer inspection. Despite the desired polycentric development at the macro-regional and national levels, where it is important to avoid polarization between capitals, metropolitan regions and other regional centres, the European perspective also put a strong emphasis on the regional dimension and the role of small and medium-sized towns in local development (EU Ministers responsible for Spatial Development, 2011). Local towns are considered to be crucial parts of less urbanized and rural areas because of their irreplaceable role as service and labour centres which ensure social and economic cohesion at the lowest geographical level. Moreover, local urban systems and networks of regional centres also have the potential to counterbalance the dominance of major cities and to contribute in consequence to a more balanced socio-economic development (RePUS, 2007). With regard to different geographical scales, the dynamics and form of interconnectedness of centres vary widely. In the context of polycentricity and disparity measures, the issue of scale must be taken into account first because it significantly affects the research design and interpretation of results (see e.g. Taylor et al., 2008; Vasanen, 2013). In the case of PUR, we can expect different forms of everyday interactions among centres than in the case of polycentric system at the national or international level. With increasing scale, the functional relations are transformed and based on different principles, for example, the daily movement of population in the case of a metropolitan area will not play such a significant role in transnational networking of cities, which is rather based on the exchange of information and institutional cooperation. The interactions between cities can take many different forms (e.g. commuting and inter-firm trades - to learn more about multiplexity of urban networks see Burger et al., 2014) whose intensity is strongly dependent on the geographical scale (van der Knaap, 2007). Apart from the European normative standpoint (see CEC, 1999; EU Ministers responsible for Spatial Development, 2011), the importance of studying polycentricity at a micro-regional scale is closely connected to processes that shape the form and structure of local urban systems. Daily-based human activities mostly take place at the micro-regional level, where spatial and functional patterns of society take the form of everyday routine (Limtanakool et al., 2009). Although even today the framework of everyday life may be confined to one city (especially for those who live and work in it), the city itself is increasingly integrated into the urban networks. Strong ties among municipalities across space are thus the characteristic feature of urban systems. Interactions can be, for example, in the form of work commuting, shopping trips or trade (e.g. retailing, intermediate goods and business services). At this scale of human activities, we can expect the emergence of PURs, which differ from other regions by their internal organization of socio-economic activities (Parr, 2004).

3 Methods and Data A methodical approach consists of four basic stages presented in a consequential order. The first step is the definition of territorial units used for an empirical analysis, the second is the evaluation of polycentricity of individual regions, the third involves assessing the level of intra-

102 regional socio-economic disparities and the final step is to measure the significance of the relationship between polycentricity and intra-regional disparities.

3.1 Delimitation of Territorial Units When assessing the level of polycentricity and intra-regional disparities, it is important to determine the geographical scale and form of regions first. The selection of the micro-regional scale is based on the three main reasons mentioned above: importance of small and medium- sized towns in local development, potential of micro-regions to generate PURs and the absence of a study (with respect to the present topic) aimed at this geographical scale (although a highly pragmatic reason, in this context it is the most relevant). Since the administrative boundaries do not often correspond to the actual spatial functioning of the territory (at least as regards the Czech space - see Halás et al., 2010; Klapka et al., 2014), the form of functional regions is used in this study. According to the method of Coombes (2002), defining Travel-to-work Areas and using a special algorithm (Mulíček and Kozel, 2012) allowing regionalization based on any input relational data (data quantifying the strength or intensity of the relationship between two spatial units), 134 Czech functional regions based on work commuting have been delimitated (see Figure 1).

Fig. 1. The Czech functional regions based on work commuting flows in 2011. Source: Data collected from Czech Statistical Office (2014a); author’s processing.

103

The inductive character of regionalization does not require a predefined fixed set of centres4 that would determine the final form of regions to some extent. On the contrary, the principle is the functional self-containment of a region, whose internal structure allows the formation of a polycentric character. The interaction is represented by work commuting flows in 2011 as travelling to work represents most of all daily trips at the micro-regional scale (Clark and Kuijpers-Linde, 1994; van der Laan, 1998) and best available data on commuting.

3.2 Assessing Morphological and Functional Polycentricity With regard to the aim of the present study, it is essential to evaluate the structure of the urban system in particular regions. The basic territorial unit for assessing polycentricity is therefore the municipality, which may or may not play a role of a centre. Assessment of polycentricity is twofold, reflecting two basic approaches, morphological and functional. Morphological dimension of polycentricity refers to a more balanced size and territorial distribution of the urban centres across the territory (Kloosterman and Lambregts, 2001; Parr, 2004). The functional (relational) dimension of polycentricity concerns the functional connections between municipalities, whereas a balanced and multidirectional set of relations is considered to be more polycentric (Burger and Meijers, 2012). As demonstrated by Burger and Meijers (2012), seemingly different approaches may be linked together even if they show different results (a morphologically polycentric region is not necessarily functionally polycentric and vice versa). Nevertheless, the methods used in this paper do not allow simple explanation of differences between both approaches since they are not part of the same formula and rather point to a more precise identification of morphological/functional polycentricity. Morphological polycentricity is evaluated using the absolute size of the centre expressed in the number of jobs5, specifically by the rank-size distribution of the largest centres and the deviation of the largest city from the regression line of rank-size distribution (primacy). Although morphological polycentricity is measured in terms of employment (as in the case of e.g. CRR MU, 2009; Burger and Meijers, 2012), it is not the only way to assess it. Alternative measures include, for example, rank-size distribution by GDP (for larger territorial units, see ESPON 2005) and population size (Meijers, 2008), or different methods reflecting other features than distribution of cities, for example, clustering of cities, characteristics of separation and spacing of cities, or connectivity (Parr, 2004; ESPON, 2005a). However, since the functioning of defined regions is based on work commuting, evaluating the importance of centres according to their job function seems to be relevant. Furthermore, focus on the municipality level automatically excludes the use of generalized data (for example GDP) that are not available at this spatial scale.

4 Traditional deductive approach determines the centres first and then gradually other settlements that are integrated to the centre by strength ties. A limitation of this approach is the assumption of nodal character of a region and the ignorance of alternative functional and spatial arrangement of urban systems (see e.g. Nystuen and Dacey, 1961; Holmes and Haggett, 1977). 5 Number of jobs = economically active population + in-commuters – out-commuters (data collected from Czech Statistical Office, 2014a).

104

Rank-size distribution provides information on urban hierarchy which is an essential indicator of mono/polycentricity6. Rank-size order of the largest centres may indicate not only an urban system with one dominant centre, a situation rather monocentric, but also an urban system that lacks strong hierarchy and is characterized by similar-sized cities, a situation rather polycentric. In fact, identification of strictly mono/polycentric systems is almost impossible, and more accurate results are provided by scores on a scale ranging from rather monocentric to polycentric systems. While a region can be considered as more polycentric when the slope of the regression line is flatter, with increasing steepness of the regression line, the urban system tends to be more monocentric. Considering the number of centres to be measured by rank-size distribution a fixed number of cities have been preferred for the reasons summarized by Meijers (2008). Finally, the sample size of 5 largest cities has been chosen (there has been a strong correlation between measures using 5 and 10 cities). Scores of the slope of the regression line and primacy measures have been standardized by the Z-score method. According to Meijers and Sandberg (2006), the resulting values have been transformed into the overall indicator of morphological polycentricity. The Z-score of 0 has been given the value of 100 and the standard deviation of 1 equals the value of 20 (e.g. 1 = 120, 0.5 = 90). The average of the values of rank- size distribution and primacy is the overall indicator of morphological polycentricity. Higher values of the indicator indicate more polycentric urban systems (on the contrary, lower values indicate more monocentric urban systems). Polycentric development should lead to a balanced spatial organization of human activities, which is within the polycentric urban systems expressed by intense mutual relations between centres (or between centres and their hinterlands). Whereas a functionally monocentric urban system is characterized by hierarchical interactions, a functionally polycentric urban system can be recognized by multidirectional flows and a higher level of reciprocal (mutual) interactions between municipalities. Functional polycentricity is thus measured by the level of reciprocal flows using data on work commuting and methods from the POLYREG project7 (CRR MU, 2009). The reciprocal component of interaction between two municipalities is the sum of a number of mutual commuters and its share in the total number of commuters: for example, from municipality A 80 commuters heading to municipality B and from municipality B 120 commuters heading to municipality A. The total number of commuters in this case is 200, the reciprocal component includes 80 commuters from A to B and 80 commuters in the opposite direction and the hierarchical component comprises 40 commuters. The sum of reciprocal commuters is then 160 and the share in the total number of commuters is 80%. This approach is applied to all the interactions between municipalities within each functional micro- region. The total number of internal commuters and the sum of reciprocal interactions are calculated for each micro-region. The share of total reciprocal component in the total number of commuters compared with the other regions indicates the level of functional polycentricity.

6 Balanced spatial distribution of cities has not been taken into account as a considerable number of regions are small spatial units where distribution of cities across the territory is not a significant factor. Presumption of mutual daily commute distances between centres is established by functional regionalization. Moreover, when assessing polycentricity at the micro-regional scale, the clustering of cities is a more important factor than balanced spatial distribution of cities (Parr, 2004). 7 The POLYREG project aimed at research on polycentric regional development in the Czech Republic, supported by the Ministry of Regional Development (under programme ‘Research into Solutions to Regional Disparities’).

105

As Burger and Meijers (2012) point out, a functional polycentric system represents centres that are relatively equal in terms of their connectivity to other centres and it is not a networked system characterized by functionally interdependent centres (higher level of network density). The level of spatial and functional interdependence is the extent to which the centres are functionally related. In this context, measuring the share of reciprocal flows illustrates the balance of functional connections between centres and does not reflect the strength of functional interdependence (network density). As in the case of morphological polycentricity, the values are transformed into the overall indicator of functional polycentricity.

3.3 Measuring Intra-regional Disparities Intra-regional disparities indicate the difference between the levels of socio-economic development of municipalities. Selection of indicators has been limited due to the availability of data at the local (municipal) level. Despite this limitation, five indicators have been selected in order to characterize the economic, social and demographic levels of each municipality. The unemployment rate and the number of active firms per inhabitant indicate the economic level, the share of university-educated people in the total population over 15 years points to social aspects, the demographic situation of municipalities is represented by age index and the crude net migration rate denotes the attractiveness of places for living. Listed categorization of economic, social and demographic indicators does not have to be so strict and should serve rather as a formal framing of selected indicators (e.g. the unemployment rate certainly indicates not only the economic performance, but also the social aspects of development). Indicators are calculated for the year 2011; only the crude net migration rate represents the overall value of the time period from 2001 to 2011. A considerable number of methods for measuring regional disparities exist (see e.g. Shankar and Shah, 2003; Webber and White, 2003; Portnov and Felsenstein, 2005). In the present study, intra-regional disparities of each region have been measured by the most common and traditional statistical measures of variability, the standard deviation and the Gini coefficient. Standard deviation is a measure of dispersion around the average and is influenced by extreme and mean values. The higher the standard deviation, the greater the intra-regional disparities. Coefficients that are independent of a mean provide a more precise measure and easier interpretation of results. Thus, the Gini coefficient is used for measuring the relative extent of regional disparities. Probably the most widely used regional inequality measure in social sciences is calculated as the arithmetic average of the absolute value of differences between all municipalities, divided by the average value of an indicator8 (Shankar and Shah, 2003).

8 The Gini coefficient normalization by Raffinetti et al. (2014) has been used for measuring the crude net migration rate. In contrast to the standard Gini coefficient, the normalized Gini coefficient is also defined for negative values and negative average. Thus, in the case of net migration (where distribution involves also negative values) the normalization is necessary.

106

The Gini coefficient (G) is calculated as follows:

1 1 G = 𝑛𝑛 𝑛𝑛 | | 2 ( 1) � � � � 𝑦𝑦𝑖𝑖 − 𝑦𝑦𝑗𝑗 𝑦𝑦 𝑛𝑛 𝑛𝑛 − 𝑖𝑖 𝑗𝑗 The arithmetic mean of the specific indicator is y, yi and yj are the values of the specific indicator for municipalities i and j, respectively and n is the number of municipalities in a particular region. If the value of the indicator is the same in all municipalities, then G = 0. For perfect inequality G = 1. With increasing value of the Gini coefficient, the disparities are greater.

3.4 Correlation of Variables Resulting values of variability measures of each region are finally compared to the level of mono/polycentricity of particular urban systems. The relationship between intra-regional disparities and the level of mono/polycentricity is evaluated using the Pearson correlation coefficient.

4 Results of the Empirical Study: Polycentricity and Disparity Levels in the Czech Regions 4.1 Levels of Polycentricity From 134 functional regions, three regions with less than five municipalities have been excluded (in order to measure the levels of polycentricity by rank-size distribution of five largest centres). Yet, the size of the regions is highly differentiated. While more than 2 million inhabitants live in the functional region of Prague (the largest Czech region), the smallest regions have only slightly more than 6,000 inhabitants (see Figure 1). Except for methodological parameters, the significant differences in the size of the regions are due to urban system hierarchy and the ability of municipalities to play the role of the centre. The results are therefore divided into categories based on the size of the regions. Table 1 shows the population distribution and number of municipalities in each size category. With increasing population size of the regions, the number of regions is smaller (with one exception in the smallest category). The uneven distribution of economic activities is documented by the three largest regions (Prague, Brno and Ostrava as the biggest centres), where almost a third of the total Czech population is concentrated. When comparing the level of polycentricity and the size of the region, both approaches to polycentricity (morphological and functional) should be assessed separately. Although a certain level of interdependence between the two polycentricity approaches is evident (see Figure 2), in fact, a morphologically polycentric region does not have to be a functionally polycentric region (and vice versa). A significant negative relationship is present between morphological polycentricity and size of the regions (see Table 1). Within the particular size categories only the morphological polycentricity of the largest regions correlates significantly and again negatively with the size of the regions. Thus, it can be suggested that more morphologically polycentric regions emerge predominantly in smaller regions in terms of population size. The level of functional polycentricity is not closely related to the population size of the regions.

107

Tab. 1. Size of the Czech functional regions and Pearson correlation coefficient between population size of the regions and levels of morphological (MP) and functional (FP) polycentricity. Correlation of population Population Number Arithmetic mean size of the regions with Size category (2011) of regions levels of polycentricity MP FP MP FP 20,000> 426,625 32 104.2 108.6 0.122 0.106 20,001-50,000 1,659,541 49 99.8 96.2 -0.057 0.030 50,001-100,000 1,931,738 27 100.7 97.2 -0.121 0.259 100,001< 6,397,259 23 93.8 99.5 -0.493* -0.270 All regions 10,415,163 131 100.0 100.0 -0.283** -0.120 * significant correlation at the 0.05 level, ** significant correlation at the 0.01 level Source: Czech Statistical Office (2014a); author’s calculation

Fig. 2. The levels of morphological and functional polycentricity in the Czech functional regions. Source: Data collected from Czech Statistical Office (2014a); author’s processing

The levels of both morphological and functional polycentricity in the Czech functional regions are illustrated in Figure 2. Regions with the highest scores of morphological and functional polycentricity are situated mostly close to the national borders with a rather random order. A noticeable concentration of these regions can be found in the north-east of the Czech Republic. One exception is the region with the largest centre, Tábor, which extends to the central part of

108 the country and can be also distinguished from other polycentric regions by its population size and area. The most morphological and functional monocentric urban systems can be observed in regions of the largest cities (Prague, Brno, Plzeň, České Budějovice and Pardubice) and in medium-sized regions (from 20,000 to 100,000 inhabitants) localized in the eastern part of the country. However, the structure of urban systems of the rest of the Czech regions cannot be clearly defined since the morphological and functional levels of polycentricity differ to a greater extent. For example, functionally polycentric regions are evidently concentrated in the north- western part of the country, where also morphologically monocentric regions are situated.

4.2 Relationship between Polycentricity and Disparities The urban system structure of each region, in terms of polycentricity level, has been finally correlated with the level of intra-regional disparities (measured by standard deviation and the Gini coefficient). Positive correlation means that with increasing level of polycentricity the intra-regional disparities grow. Negative correlation means the opposite, with increasing level of polycentricity the intra-regional disparities are lower (see Figure 3). Being aware of the differences between morphological and functional polycentricity, the correlation has been measured twice (for both approaches to polycentricity). In addition, the measurement has been also carried out for each size category of regions. Considering these aspects and five indicators of development levels, the resulting number of correlation values is 100. When interpreting the relationship between polycentricity levels and intra-regional disparities, it is necessary to take into account the type of polycentricity (morphological/functional), the size of the region, the chosen method for measuring disparities and the specific indicator of intra-regional disparities. The results show (see Table 2) that the connection of polycentricity to the level of intra- regional disparities is not unambiguous. A significant relationship is mostly formed when evaluating all regions. This is partly a consequence of inner differentiation in the population size of regions. Firstly, a significant relationship for all regions is achieved despite the absence of a significant relationship within particular size categories (e.g. morphological polycentricity with age index and university-educated measured by the Gini coefficient). Secondly, a significant relationship for all regions exists because of the correlation value in specific size category, where the significant relationship substantially exceeds the values measured in other size categories (e.g. morphological polycentricity with university-educated and crude net migration rate measured by standard deviation, and functional polycentricity with university- educated measured by the Gini coefficient). In this case, the significant relationship for all regions is the result of only one (or two) significant correlation(s) which belongs to one size category. Besides the impact of population size categories of regions on results, different levels of morphological and functional polycentricity are also reflected in various correlation values. One possible consequence is that in case of a relationship between morphological polycentricity and a selected indicator of intra-regional disparities, the correlation is significant, but in the case of functional polycentricity there is no significant relationship (and vice versa). However, the situation of opposite significant relationships (positive and negative) related to different levels of morphological and functional polycentricity does not occur.

109

Fig. 3. Examples of positive and negative correlations between levels of morphological polycentricity and the specific indicator of intra-regional disparities measured by the Gini coefficient (G) and standard deviation (SD). Source: Data collected from Czech Statistical Office (2014a, 2014b); author’s processing

Finally, correlation values depend on the chosen method for evaluating intra-regional disparities. The difference between standard deviation and the Gini coefficient is primarily visible when focusing on measures for all regions and the correlation of morphological polycentricity with the crude net migration rate. While a significant positive correlation results from the measure by the Gini coefficient, a significant negative correlation occurs when standard deviation is used. Differences of these two measures are caused not only by their methods of calculation but also by specific distribution of the crude net migration values. Crude net migration rate is characterized by greater disparities and extreme values than other indicators. Thus, standard deviation is more influenced by such distribution and differs significantly from Gini coefficient values. Nevertheless, the above-mentioned difference deserves a deeper explanation. Negative correlation (measured by standard deviation) reaches the highest value in the category with more than 100,000 inhabitants. Since the number of municipalities is mostly larger in more populated regions (at least in Czech regions), the probability of more extreme values is higher in these regions and the standard deviation values indicate greater intra-regional disparities. Reflecting the most populated Czech functional regions with the largest number of municipalities, the majority of them are rather monocentric with one dominant centre and are typical areas with the most intensive process of suburbanization in the observed period (e.g. regions of Prague, Brno and Plzeň). Suburbanization significantly contributes to an increase in the crude net migration rate and may widen the distribution of values. As a result, the intra- regional disparities in the crude net migration rate are greater in these monocentric regions. On the contrary, polycentric regions with more than 100,000 inhabitants are characterized by smaller number of municipalities and lower level of suburbanization (e.g. regions of Nový Jičín, Uherské Hradiště, Teplice and Frýdek-Místek). Thus, intra-regional disparities are lower in

110 these regions. Since the crude net migration rate gives us a picture about the attractiveness of a particular place it can be suggested that the most populated and monocentric regions tend to be fragmented in local attractiveness for living. The significant positive correlation of morphological polycentricity and intra-regional disparities in the crude net migration rate measured by the Gini coefficient for all regions originates in the smallest population size category (less than 20,000 inhabitants). The intra-regional disparities in the crude net migration rate increase with higher levels of morphological polycentricity. Since the Gini coefficient measures inner structure of distribution of values (differences between all pairs of values) it provides a more precise view of the relative extent of intra-regional disparities. Concerning the correlation of polycentricity with socio-demographic indicators, age index and university-educated people in the total population over 15 years, both of them show similar results. With the increasing level of polycentricity, the intra-regional disparities in age index and university-educated people in the total population over 15 years decrease. The concentration of university-educated population in the largest centres with sufficient number of jobs is a common trend that holds true also in the Czech Republic (based on own calculation of data from Czech Statistical Office, 2014a). Accordingly, polycentric regions provide more places that are preferred by people with higher education. A more even distribution of university-educated population can be observed predominantly in polycentric regions with more than 100,000 inhabitants (e.g. regions of Teplice and Uherské Hradiště). On the other hand, excessive concentration of people with university education can be found in strongly monocentric regions (e.g. region of Prague). The largest centres of each region attract university-educated people as well as other young people who are looking for employment opportunities. Young immigrants very often start families in these centres and the share of the youngest (or pre-productive) population is rising. However, the processes of suburbanization and urban shrinkage (connected to post-industrial transformation of former industrial centres) could modify this trend. In the case of Czech regions, age index is more evenly spread in polycentric urban systems and more monocentric regions are characterized by a rather unequal demographic structure in terms of distribution of population by age. Intra-regional disparities in economic performance are represented by unemployment rate and active firms per inhabitant. The relationship between economic disparities and polycentricity has not been proved. However, two exceptions exist. A significant negative correlation of functional polycentricity with intra-regional disparities in unemployment rate (measured by the Gini coefficient) in all regions indicates the tendency of unemployment rate to decrease in more polycentric regions. A positive correlation emerges between functional polycentricity and disparities in active firms per inhabitant measured by standard deviation. This relationship concerns regions with 50,000-100,000 inhabitants. In general, intra-regional disparities in active firms per inhabitant are higher in more polycentric regions. The random character of significant relationships does not provide a solid basis to more specific conclusion. Despite these significant correlations, it can be concluded that a provable relationship between disparities in economic performance and levels of polycentricity has not been found. When interpreting the relationship between structure of urban systems and intra- regional disparities described above, the other historical, cultural or social factors should be taken into consideration (but cannot be inferred from the analysis). Although spatial

111 organization of society has changed in recent years, particular urban systems may have (not) the same structure even after the economic and social transformation (for example deindustrialization and consequent tertiarization of society) or progress in technologies, communication services and transport infrastructure, no matter how economic or social performance of the region has changed. The original spatial structure of a specific urban system (for example morphologically polycentric) could persist despite the restructuring of the economy because of the stability of the urban system and adaptive skills of the local workforce. A particular urban system may thus maintain its shape but lose its economic and social standard. These arising thoughts, however, concern the level of economic and social development which cannot be expressed by disparity measures.

112

Tab. 2. Pearson correlation coefficient between polycentricity levels and disparity measures. Size category

Disparity measure 20,000> 20,001-50,000 50,001-100,000 100,001< All regions

MP FP MP FP MP FP MP FP MP FP unemployment rate SD 0.139 -0.066 -0.198 0.013 -0.322 -0.088 -0.126 0.047 -0.099 -0.045 G 0.038 -0.283 -0.167 -0.266 0.039 0.059 -0.276 -0.234 -0.135 -0.220* active firms per one SD 0.049 -0.055 0.104 -0.007 -0.115 0.429* -0.406 0.062 -0.035 0.047 inhabitant G -0.023 0.049 -0.033 0.008 -0.176 0.346 -0.073 0.391 -0.031 -0.160 university-educated in total SD -0.190 -0.255 -0.302* -0.202 -0.112 -0.126 -0.661** -0.597** -0.350** -0.278** population over 15 years G -0.112 -0.246 -0.203 -0.204 0.060 -0.301 -0.355 -0.110 -0.200* -0.030 age index SD -0.017 -0.212 -0.189 -0.307* -0.363 -0.357 -0.165 -0.212 -0.138 -0.239** G -0.140 -0.298 -0.204 -0.228 -0.361 -0.166 -0.222 -0.121 -0.199* -0.235** crude net migration rate SD 0.052 -0.079 -0.065 0.210 -0.301 0.302 -0.642** -0.114 -0.250** 0.026 G 0.645** -0.120 0.241 -0.029 -0.165 -0.283 0.318 -0.073 0.326** -0.050 * significant correlation at the 0.05 level, ** significant correlation at the 0.01 level Source: Czech Statistical Office (2014a, 2014b); author’s calculation

113

5. Conclusions Globalization, post-industrial economy and development of new technologies and innovations have strengthened the role of the largest urban centres in past decades. Cities are increasingly functionally integrated into the broader urban systems and no longer act as isolated spatial units. Traditional central place systems theories presenting hierarchical models of urban systems have been modified and gradually replaced by more complex representations emphasizing urban networks and the polycentric structure of urban systems (Davoudi, 2003). Polycentric urban systems lack a clear hierarchy structure with one dominant centre and are rather characterized by multiple centres with mutual interactions (Kloosterman and Musterd, 2001). Promoting polycentric development has become a corner stone not only of EU regional policy but also of many European countries (CEC, 1999; Meijers et al., 2005). One of the presumptions is the ability of polycentricity to diminish territorial socio-economic disparities. However, the connection of polycentric urban systems to territorial disparities has not been properly tested (except for studies made by, for example, Meijers and Sandberg, 2006; Veneri and Burgalassi, 2012), in particular at the micro-regional scale. Thus, the aim of this study has been to investigate the relationship between levels of polycentricity and intra-regional disparities. Scale-dependency of the polycentricity concept complicates its definability and generates the necessity of assessing polycentricity at different scales separately. At the micro-regional scale, the role of regional and local centres is important in terms of providing basic services and jobs. Work commuting flows in 2011 have been used to delimit functionally self-contained micro-regions and both morphological and functional approaches have been applied to assess the level of polycentricity. As morphological and functional polycentricity describes different aspects of urban systems (although interrelated), levels of both polycentricity measures do not correspond. While the most morphologically monocentric regions are generated around the largest dominant centres of Prague, Brno or Plzeň, more morphologically polycentric regions can be found especially close to the national border and in the areas circling the largest functional regions. Some of the morphologically polycentric regions are furthermore polycentric in functional aspect. However, functional polycentric regions situated in the north- west of the Czech Republic can hardly be considered morphologically polycentric. In general, the largest regions are rather monocentric and for the smallest regions, the urban systems can be characterized as more polycentric. Based on this conclusion, levels of intra-regional disparities are closely linked to the population size of a micro-region. While the largest and most populated micro-regions have a strong centre (majority of cases) and the probability of greater internal disparities is higher, smaller polycentric micro-regions are situated outside the metropolitan areas and do not have a strong centre which would polarize the job function of municipalities. Consequently, in spite of the expected lower degree of intra- regional disparities within smaller polycentric micro-regions specific services and jobs are offered only by an external (behind the border of a particular micro-region) centre. On the other hand, notwithstanding the greater intra-regional disparities within the most populated micro- regions the presence of a bigger centre guarantees better accessibility to specific functions (services and jobs). Bearing this in mind, results of correlation should be interpreted carefully and with awareness of necessary generalization. The level of intra-regional disparities is based on the social, demographic and economic indicators of each municipality and compared with

114 the polycentricity level of a specific region. In the case of Czech functional micro-regions, intra- regional disparities in socio-demographic development (represented by university-educated people in the total population over 15 years and age index) decrease when the urban systems are more polycentric. Centres in each region naturally provide more qualified jobs and attract young families, while periphery localities suffer from population ageing. In spite of contradictory tendencies modifying this trend (suburbanization and counterurbanization), the distribution of the young and university-educated population is strongly fixed to the largest urban centres in a particular region. Hence, in the context of socio-demographic development, the assumption about the impact of polycentricity on lowering intra-regional disparities holds true. Such a relationship disappears if disparities in economic performance are considered. The relationship between polycentricity and economic performance has not been proved. Attractiveness of places for living has been determined using the crude net migration rate during the period from 2001 to 2011. When comparing levels of polycentricity with intra-regional disparities in attractiveness of places for living, the results are rather unclear due to two fold measurement. Here, again, the relationship of polycentricity with attractiveness of places has not been unambiguously defined. Every evaluation of disparities and measurement of the relationship between two variables brings generalization which is necessary, but rarely provides a deeper understanding of causal connections. However, the comparison of several regions is the main purpose of this study and a quantitative approach seems to be inevitable. We should realize that the availability of data is largely limited, especially at the municipality level. Moreover, measurements using slightly different data could change the results. Similarly, different methods of variability measures could offer different levels of disparities. Despite these methodological limitations, the study provides results that partially coincide with previous studies on this topic (Meijers and Sandberg, 2006; Veneri and Burgalassi, 2012; Häzners and Jirgena, 2013). Even the analysis at the micro-regional level has not proved the unique role of polycentricity as an accelerator of balanced spatial development. Nevertheless, the impact of polycentricity on equal socio-demographic spatial development is rather positive. Certain ambiguity of results raises questions about the scale and delimitation of territories and chosen methods and represents true motivation for further research. On top of this, comparison of results with that of other European countries would provide a desired examination of the polycentricity issue in a broader context.

115

116

Paper 4: Small towns in the context of ‘borrowed size’ and ‘agglomeration shadow’ debates: The case of the South Moravian Region (Czech Republic)

JIŘÍ MALÝ

European Countryside (in press)

ABSTRACT Small towns play a key role in providing services for its wider hinterland. However, emerging economic importance of the largest agglomerations and increasing involvement of settlements in urban networks have transformed a relationship between the size of settlements and their expected urban function. In this context, the concepts of ‘borrowed size’ and ‘agglomeration shadow’ serve to explain the impact of network externalities on urban function but pay a little attention to service function of small towns. The paper aims at revealing the extent to which the provision of services is determined by location of small towns within a regional urban system strongly affected by a metropolitan area. The results show coexisting occurrence of the processes of borrowed size and agglomeration shadow and the importance of tourist and commercial attractiveness of particular places (towns) to final provision of services.

Keywords: small towns, borrowed size, agglomeration shadow, service function, periphery, metropolitan area

1 Introduction The hierarchy of urban systems conceptualized by traditional Christaller’s Central place theory has been confronted by models accentuating complexity of urban networks emerging from the processes of globalization and growing mobility of contemporary society (van der Laan 1998). While the role of the largest agglomerations is emphasized by proponents of the new economic geography (Fujita and Krugman, 2004; Glaeser, 2011) and the most of the leading urban system theories are focused on densely populated areas acting as the economic growth poles in regional development, e.g. ‘mega-city regions’ (Hall and Pain, 2006; Florida et al., 2008), ‘polycentric urban regions’ (Kloosterman and Musterd, 2001; Parr, 2004), the role of small towns seems to be neglected despite their importance especially in rural areas. Until recently, attention dedicated to small and medium-sized towns appeared to be closer in studies focusing on the issue of rural development (Zekeri, 1994; Courtney et al., 2007). However, the resumption of a discussion about the concept of ‘borrowed size’ (Phelps et al., 2001; Meijers et al., 2016) contested by the ‘agglomeration shadow’ effect (Burger et al., 2015) has brought the issue of the role of small towns in regional development to wider research

117 interest including urban and economic geography. An urban function of small towns is relativized by the position within an urban network and the access to socio-economic resources. Put it simply, population size of a particular place does not have to correspond to its expected urban function due to an advantage stemming from the proximity of an urbanized area (borrowed size) or due to a competition with larger urban centre (agglomeration shadow). The most of studies deal with the (inter)national scale where the higher-ranked functions are at the forefront of interest and the basic spatial unit very often represents larger spatial entity (Partridge et al., 2009; Tervo, 2010; Burger et al., 2015; Meijers et al., 2016). Hence, this paper intends to focus on regional dimension allowing to explore the spatial urban pattern more deeply and on wide spectrum of services contributing to the functioning of small towns as urban centres with different intensity. The main aim of the paper is to evaluate the service function of small towns (on the case of the South Moravian Region, Czech Republic) with regard to their location within the urban and the transport network, their historical importance and tourist attractiveness. The hypothesis is related to the concept of agglomeration shadow and it claims that the service function of small towns is affected, next to its own population size, also by the position of the town towards the core city of a metropolitan region and the transport infrastructure. The results of this relationship are interpreted with respect to the site-specific features forming the spatio-functional dynamic of the region.

2 Theoretical background The role of small towns within urban systems has been partly overlooked by urban researches in recent scholarly debates about regional development. However, their importance for rural areas are crucial in order to maintain balanced spatial development and to avoid spatial discrepancies (Hinderink and Titus, 2002). The specific role of small towns consists of the sufficient provision of jobs and services for their mostly rural hinterlands (Heffner and Solga, 2006; Vaishar and Zapletalová, 2009). This assumption is reflected in normative practices of spatial planning authorities at different hierarchical levels. Regarding non-binding platform of the ‘European spatial planning policy’ the Territorial Agenda 2020 suggests that ‘in rural areas small and medium-sized towns play the crucial role; therefore it is important to improve the accessibility of urban centres from related rural territories to ensure the necessary availability of job opportunities and services of general interest’ (EU Ministers responsible for Spatial Development, 2011, p. 8). Improved living conditions in rural and peripheral areas related to the role of small towns has become a strong part of the territorial cohesion discourse shaping EU regional and sectoral policies in recent years (EC, 1999; EC, 2008). At the Czech national level, the highest ranked spatial planning document called Spatial Development Policy refers to small towns in the context of better access to basic infrastructure. It states that one of the priorities is to ‘create conditions for the improvement of transport accessibility of municipalities (towns) that are natural regional centres (while benefit from their location and infrastructure) in order to improve conditions for the development of surrounding municipalities localized in rural areas and areas with specific geographic features’ (MMR, 2015a, p. 15). Although the undisputable position of rural and peripheral areas in spatial development is well formulated in debates about territorial cohesion and embedded in spatial planning documents (EU, national level), recent trends characterised by increasing globalization,

118 mobility and growing importance of large agglomerations in overall development tend to enhance polarization patterns and spatial disparities (Ezcurra et al., 2005). From the perspective of balanced development, polarization between particular territories or places is seen as undesired feature of spatial development and is very often understood as a stimulus for growth of regional inequalities (EC, 2008). Molle (2007, p. 90) argues that polarization ‘implies not only the aspect of geographical concentration but also the accentuation of the differences in endowments, equipment and hence disparities in wealth between the core and the periphery’. Indeed, spatial disparities are closely linked to the social welfare and have strong impact on socio-economic situation of communities and individuals (and vice versa). As a part of prevailing neoliberal economic model of contemporary European society, complex system of socio-spatial interdependencies face challenges produced by global market forces, changing spatial divisions of labour and demographic structure affected by population aging. Increasing emergence of socio-spatial disparities among the European territory gave rise to discussion about the theoretical concepts of ‘polarization’ and ‘peripheralization’ – attempts to explain socio-spatial imbalances and their origins (Hudson, 2015; Kühn, 2015; Lang, 2015). Despite the complexity of production of peripheries, its spatial dimension is very often linked to the structure of urban system and especially to metropolitan and non-metropolitan areas and core-periphery relation (Claval, 1980; Copus, 2001; Krätke, 2007). Although urban systems are characterized by significant stability in time, recent changes in communication technologies, transportation systems and production modes resulted in a transformation of urban systems with varied impacts at different spatial scales (Castells, 1996; van der Laan, 1998). During recent decades of more intensive research on urban systems, the structure and functioning of urban systems have been described by large number of theoretical frameworks (Meijers, 2007 provides an overview), ranging from Christaller’s Central place theory (Christaller, 1933), general ‘hierarchical urban models’ (Berry & Parr, 1988), ‘network models’ (Batten, 1995; Camagni and Salone, 1993), to the concept of ‘polycentricity’ (Kloosterman and Musterd, 2001; Parr, 2004). Theoretical models based on hierarchical order and size of cities indicating their centrality have been replaced by network paradigm emphasizing interactions among cities, specialization of centres and importance of horizontal linkages in terms of human activities (Capello, 2000). Morphological approaches have been enriched with functional dimension of urban system’s structure (Green, 2007). In this context, position of places within horizontal linkages creating integrated network of cities and their functional connection to metropolitan areas have become crucial aspects of economic development and potential success in regional competition. In contrast with rather schematic conceptualization of urban systems from the perspective of theories emphasizing a hierarchical order and a linear connection between size and function of a particular settlement unit, the models operating with the concepts of network or polycentricity provide a basis for contesting the linkage between the size of settlement and its function within an urban system. As Batten (1995) argues, the position of a particular place within an urban network is crucial for its function (in terms of e.g. economic or cultural importance). Not only the own size but also the size of its hinterland affects the centrality of a particular place. Consequently, ‘due to the presence of spatial interdependencies, smaller places can borrow size and host functions that they could not have hosted in isolation’ (Burger et al.,

119

2015, p. 1092). Although the concept of borrowed size is not new in geographic and economic research, only a few studies reflecting its impact on urban systems have been conducted (Alonso, 1973; Phelps et al., 2001; Meijers et al., 2016). While even small towns may benefit from higher population density beyond their administrative borders and therefore attract people (e.g. workers, consumers) from wider urban network, the effect on the function of a particular town could be exactly opposite. In this respect, the concept of agglomeration shadow aims at explanation of fewer functions that are provided by a town located in a proximity to larger centre compared to a town isolated from other urbanized parts of a given territory (Fujita et al., 1999; Burger et al., 2015). In the context of borrowed size versus agglomeration shadow discussion, metropolitan processes including residential and commercial suburbanization become highly relevant. Studies researching agglomerations and metropolitan zones are mostly focused on social and spatial changes in the service (shopping) function of hinterland (Heffner and Twardzik, 2013; Maryáš et al., 2014) or travel-to-work behaviour (Schwanen, 2002). At the international level, Burger et al. (2015) have explored the linkage between an urban function of cities (provision of high-end cultural amenities) and their (inter)national accessibility. They conclude that places lying in a shadow of larger cities are characterized by lower levels of high-end cultural amenities than could be expected considering their size. However, borrowing size of smaller cities from the larger ones has not been proofed. Hence, in general, ‘central place’ logic still persists, at least in the case of high-end cultural amenities in North-West Europe. On the example of the American urban system, Partridge et al. (2009, p. 461) suggest that ‘rural counties and smaller urban centres have significant positive interactions with their nearest higher–tiered urban areas’. Deconcentration tendencies supporting urban sprawl significantly contribute to intensive commuting linkages between urban cores and peripheries. In this context, lower–ordered places benefit from closer accessibility of urban amenities in terms of population dynamics (Partridge et al., 2009). While provision of high-end cultural amenities is subject to agglomeration shadow effect, population growth of places situated in the hinterland of urban cores may persist. In fact, borrowing size versus agglomeration shadow effects are strongly dependent on a type of amenities and other aspects (population growth, housing projects) used in an analysis as the urban function indicator. Moreover, a proximity to larger agglomerations should be complemented by other factors (such as connection to transport infrastructure, historical aspects, cultural differences or tourist attractiveness) that may have an impact on the overall urban function of a particular place. In this context, Knox and Mayer (2012) see the comparative advantage of small towns specifically in their cultural heritage, potential for liveability and sustainability, and attractiveness for leisure. Additionally, the issue of geographical scale predetermines the scope and type of analysed functions and therefore the final conclusions about agglomeration shadow intensity. Thus, the next parts of the paper aims at the regional scale, where the service function of small towns is evaluated reflecting not only agglomeration shadow and borrowing size debate but also other factors that could play a role in the level of an urban centrality. The next chapter presents the case study area and the research methods, the empirical results are presented in the following chapter, and the main conclusions are summarized at the end of the paper.

120

3 Study area, methodology and data The South Moravian Region situated in the south-eastern part of the Czech Republic has been chosen as a case study area mainly due to its monocentric urban system structure that is strongly influenced by an agglomeration of the regional capital Brno (Malý, 2016). The rest of the region is composed predominantly by a large number of small and medium-sized towns that create scattered urban system with high density of rather small municipalities (see Figure 1). The methodology consists of the four main steps. Firstly, the services and their values used to calculation of the service function have to be defined. Secondly, specification of parameters defining small towns is essential for their selection and calculation of their service function centrality. Thirdly, the variables with potential impact on service function centrality of small towns are determined. Finally, a relationship between the variables and the scores of service function centrality of small towns is evaluated using regression analysis.

Fig. 1. The urban system of the South Moravian Region. Source: Czech Statistical Office (2014c), BMA delimitation by Mulíček et al. (2013); author’s processing

The final list of 30 basic and higher-ranked services together with the number of facilities localized within the South Moravian Region territory (calculation for 672 municipalities) has been used to design service’s value. In line with the central place theory a distribution of services is not spatially balanced but is rather hierarchical respecting the urban system structure (see Table 1). While the most common services, e.g. public library, grocery shop, and pub, are characterized by the largest number of facilities and are provided by even small municipalities, services with a limited number of facilities are concentrated into the largest cities.

121

Tab. 1. Distribution of services among the municipalities in the South Moravian Region. Number of Average size of Services Service’s municipalities with municipality with (2012-2014) value particular facility particular facility Public library 628 1,855 1 Grocery shop 605 1,917 1 Pub 592 1,956 1 CzechPoint a 591 1,952 1 Cultural centre 563 2,009 1 Early childhood education 423 2,626 5 Primary education 334 3,167 5 Restaurant 310 3,264 5 Post office 293 3,482 5 Bank 280 3,618 5 General practitioner 210 4,508 10 Paediatrician 161 5,582 10 Low-secondary education 153 5,814 10 Dentist 140 6,308 10 Drugstore 109 7,656 10 Museum and gallery 71 10,031 15 Nursing home 71 10,172 15 Cinema hall 65 11,233 15 ATM 61 12,320 15 Insurance company 53 13,644 15 Upper secondary education 31 20,266 20 Retirement home 29 19,937 20 Grammar school 21 28,516 20 Emergency medical service 20 29,892 20 Hospital 12 46,186 20 Facility for physically disabled people 12 35,638 20 Shelter 9 57,912 25 Low-threshold facility for children and youth 8 60,719 25 Shopping centre 7 68,047 25 Theatre 4 108,350 25 a assisted place of public administration Source: The South Moravian Region (2012), Czech Statistical Office (2013), Banky (2015), Czech Post (2015), CzechPoint (2015); author’s calculation

A hierarchical order is also reflected in a transportation accessibility of particular services (see Figure 2). A comparison of the access to the nearest grocery shop, hospital and theatre indicates the differences based on the range of services. The lowest threshold of grocery shops is related to smaller distances (range) that are needed to reach the service (the shortest time and the largest number of public transport connections). On the contrary, the largest distances of theatres reflects higher threshold and specialty of the service. The service’s value has been determined according to the number of municipalities with particular facility (the first numeric column in Table 1 – over 500 = service value 1; 250-499 = 5; 100-249 = 10; 50-99 = 15; 10-49 = 20; less

122 than 10 = 25). Such rating aims at capturing the threshold and range of services in a particular territory. The final index of service function centrality for each municipality is based on the service’s value and is calculated as the sum of the number of facilities of each service provided by a particular municipality multiplied by the service’s value1. The index of service function centrality is calculated for each municipality.

Fig. 2. Transportation accessibility of selected services – the nearest facility of a particular service for each municipality in the South Moravian Region measured by the car time distance (left) and the number of public transport connections during a week-day (right) and sorted from a municipality with the highest accessibility (1) to a municipality with the lowest accessibility (672). Source: The South Moravian Region (2012), Czech Statistical Office (2013), IDOS (2015); author’s calculation

Definition of small towns is the crucial issue and it is strongly dependent on a particular urban system hierarchy. In the case of the South Moravian Region, the maximum population size is set to 15,000 inhabitants (besides the metropolis of Brno with almost 400,000 inhabitants, there are five settlements with population size between 20,000 and 35,000 that are considered, together with Brno, large centres) which is in accordance with the studies made by Cigale et al. (2006) and Vaishar et al. (2015). However, the minimum population size is not defined. The main reason is to consider the possible effect of borrowed size issue in the case of a relatively numerous group of municipalities with approximately 3,000 (and less) inhabitants and to examine the differences in service function centrality of the smallest centres. Thus, the lower limit is based on the index of service function centrality – the value has to be 80 at minimum. The rest of municipalities (with the value lower than 80) are predominantly small villages with almost no service function and including them into analysis does not change the results. The final set of municipalities will be termed ‘small centres’ because of relatively considerable number of municipalities without the ‘town’ status serving as centres for their hinterland and providing urban functions. Moreover, the status of a ‘town’ is one of the independent variables

1 For example, the index of service function centrality for a municipality providing 2 pubs, 1 grocery shop and 2 restaurants is calculated as follows: (2 x 1) + (1 x 1) + (2 x 5) = 13.

123 entering regression analysis and therefore labelling the selected municipalities ‘towns’ would be confusing. The index of service function centrality of the final list of small centres is then described according to population size categories and analysed through the relative values of the index (since the population size of small centres varies to a great extent, service function centrality is relativized by the number of inhabitants in order to conduct more efficient comparison of data). The next step is to define variables that could have an impact on the service function of small centres. Regarding a discussion about the concepts of borrowed size and agglomeration shadow the characteristics reflect spatial position of small centres within the urban system and interaction with higher-ranked centres (area: Position within the urban system); transport infrastructure and connection of small centres to other urbanized parts of the territory (area: Access to transport infrastructure); as well as their tourist attractiveness and historical importance (area: Inner potential) that also might be the factors contributing to different levels of service function centrality (see Table 2). Particular variables are set up for each of these areas capturing the main characteristics that fit into a particular area of interest. As regards inner potential, the historical importance is defined by two variables – town status and the role of small centre during the period of socialist central planning. While the town status stems from a significant role of a particular centre in the deeper history, the socialist period had an impact on urban system hierarchy especially by defining centres for rather planning and ideological purposes (Malý and Mulíček, 2016). Tourist attractiveness is measured by the number of accommodation facilities. Position of small centres within the urban system include, besides the evaluation of rural character of centres and their peripherality, also the importance of small centres at higher geographical scale by measuring their interaction with large centres in the region. While the existence of out-commuting flows to higher-ranked centres are common to most of small centres, in-coming flows of work commuters from large centres to smaller municipalities in their hinterland present more sporadic aspect of spatio- functional dynamic of territories. In this sense, share of work commuters from large centres in the total number of employees in a particular small centre is the variable assessing the importance of small centre within the network of large centres. The level of crude net migration rate is examined in order to explore whether the increase of population size (suburbs) have a positive or negative impact on the service function. The access to transport infrastructure is based on the accessibility of motorways and public transport, the characteristics of functional polycentricity and network density reflect plurality of choices and the overall provision of public transport connections respectively (preconditions for borrowing size). The mutual relationship between spatial characteristics and service function centrality of small centres is evaluated by using a multinomial logistic regression which is able to predict a dependent variable given one or more independent variables. The relative values of the index of service function centrality were classified into three categories (high centrality, medium centrality, low centrality) which serve as the dependent variable. The variables focusing on spatial characteristics of small centres are predictor (or independent) variables. The main aim is to better understand the relationship between spatio-functional context and service function centrality of small centres. In other words, regression analysis can provide an important information if there are variables that increase the probability of having higher service function

124 centrality than other similar-sized centres. Variables were checked and corrected in order to meet all assumptions that are required for multinomial logistic regression to give us a valid result where it was necessary. Regression analysis is conducted for the entire region and Brno metropolitan area (BMA) separately in order to reflect stronger interactions and specific dynamic of the closest hinterland of the regional capital Brno. The main findings are presented in the next chapter.

Tab. 2. Distribution of services among the municipalities in the South Moravian Region. Areas of interest Characteristics and variables . Historical importance - Town status Inner potential - Centre delimitated by socialist central planning . Tourist attractiveness - Accommodation facilities (per 1,000 inhabitants)

. Rural character - Employment in agriculture (% of total employment) . Peripherality - Population density of hinterland (the area accessible in 20 minutes ) Position within - Car time distance to the nearest large centre the urban system . Importance within the network of higher-ranked centres - Share of work commuters from large centres in the total number of employees in a municipality . Suburban character - Crude net migration rate (period 1990-2015)

. Access to the nearest motorway - Car time distance to the nearest motorway (entrance) . Functional polycentricity a Access to - Number of public transport connections to the three large centres transport that are best connected – the coefficient of variation as an indicator

infrastructure of mono/polycentric structure . Network density a - Number of public transport connections to the three large centres that are best connected – the sum of the values as an indicator of the town’s integration into the urban network a to the date: 18 March 2015 Source: Terplan (1985), MFČR (2013), IDOS (2015), Czech Statistical Office (2014a, 2014c, 2016); author’s calculation

4 Empirical findings The final number of small centres in the South Moravian Region, which stems from the methodological procedure described above, is 113. In contrast with the rest of municipalities, these small centres may act as local or micro-regional centres since higher service function is assigned to them. In general, provision of services is characterized by a strong spatial

125 differentiation in the South Moravian Region (see Figure 3). While the urbanized parts of the territory (medium-sized towns and the city of Brno) keep higher-ranked services and thus the highest values of service function centrality, small municipalities situated mostly in western rural part of the region are distinguished by very low values of service function centrality. In these rural areas, small centres are very important places that provide basic services for their hinterland but also the places that are not able to satisfy human needs when higher-order services (specific culture, social, and health care facilities) are taken into account.

Fig. 3. The index of service function centrality for municipalities in the South Moravian Region. Source: The South Moravian Region (2012), Czech Statistical Office (2013), Banky (2015), Czech Post (2015), CzechPoint (2015); author’s processing

Small centres in the South Moravian Region represent only marginal part of the territory if the number of municipalities is considered (approximately one sixth of municipalities). The vast majority of settlements are small municipalities with less than 3,000 inhabitants. Besides BMA, mostly rural municipalities are functionally dependent on local centres. In general, the average absolute value of the index of service function centrality of small centres is greater when the population size increases (as central place theory suggests). However, if centres are to be compared by using the relative values of service function centrality, different results will be gained. In other words, relativizing the absolute value by the number of inhabitants allows us to mutually compare the values of the index between municipalities (belonging to small centres) with different size. The relative values give us a clearer picture of actual service function centrality since they are able to say that, for example, service function centrality of a centre A (in absolute terms lower centrality than centre B) is higher within the framework of similar-

126 sized municipalities, than if we have a look on a position of a centre B within a distinct group of municipalities characterized by similar size to the centre B. In this context, it seems to be efficient to classify small centres into particular population size categories in order to show that even similar-sized centres may differ by service function centrality to a great extent. Each population category is composed of centres with different levels of centrality and thus the inner variability of centrality for each category is significant (see Table 3). According to the results, the maximum value of relative service function centrality in the lowest population category (30.6) is even greater than the maximum value within the ‘largest’ of small centres (10.3). This is in accordance with a previous example about centres A and B. Together with the information provided by the coefficient of variation we can observe that each population category of small centres is characterized by the presence of centres providing diverse spectrum of services and different number of particular service facilities.

Tab. 3. The index of service function centrality for small centres in the South Moravian Region. Municipalities The index of service function centrality Number Average Coefficient of Population category Max. Min. Variation (2015) abs. rel. (%) abs. rel. a (rel.) (rel.) (rel.) 370,000 < 1 0.1 - - - - - 20,000 - 35,000 5 0.7 - - - - - 9,000 - 11,999 5 0.7 877 7.9 10.3 5.5 27.0

s 6,000 - 8,999 7 1.0 549 7.6 11.0 4.6 31.3 4,000 - 5,999 12 1.8 386 7.5 11.9 3.9 36.5

centre 3,000 - 3,999 17 2.5 213 6.0 12.1 2.4 44.6 2,000 - 2,999 31 4.6 154 6.2 12.9 3.7 34.3

Small 1,000 - 1,999 29 4.3 111 8.0 16.2 4.8 35.4 600 - 999 12 1.8 109 14.2 30.6 9.7 38.8 non-central 559 83.2 - - - - - a the index of service function centrality per 100 inhabitants Source: The South Moravian Region (2012), Czech Statistical Office (2013, 2014c), Banky (2015), Czech Post (2015), CzechPoint (2015); author’s calculation

Based on the inner variability of service function centrality of similar-sized centres we can state that the population size can serve only as a general indicator of the urban function and it is unable to explain differences regarding towns of comparable population size. The relative values and variation measures give us a reason to perform a regression analysis which is able to evaluate association between the service function variability and independent variables (spatial characteristics). In the case of all small centres, there are couple of factors affecting the levels of service function centrality. As regards logistic regression model for medium service function centrality (see Table 4), small centres with stronger tourist attractiveness situated rather in peripheral areas in terms of low population density and greater distance to the nearest large centre are more likely to keep the medium centrality than the low.

127

Tab. 4. Logistic regression model for the factors (independent variables) affected service function centrality of small centres in the South Moravian Region and BMA. All small centres BMA small centres a Independent variables Exp(B) Sig. Exp(B) Sig. Medium centrality Historical importance - town status 2.278 0.304 - - Historical importance - socialist centre 0.617 0.571 - - Tourist attractiveness 2.202 0.007 1.253 0.199 Rural character 1.007 0.942 - - Peripherality - density 0.579 0.038 1.114 0.623 Peripherality - distance 1.190 0.011 1.270 0.024 Importance for higher-ranked centres 1.530 0.178 2.190 0.000 Suburban character (net migration) 1.504 0.122 0.678 0.016 Access to the nearest motorway 1.039 0.186 0.993 0.893 Functional polycentricity 0.887 0.582 1.152 0.305 Network density 1.237 0.456 1.470 0.038

High centrality Historical importance - town status 2.996 0.258 - - Historical importance - socialist centre 0.804 0.829 - - Tourist attractiveness 3.075 0.000 2.159 0.001 Rural character 0.999 0.994 - - Peripherality - density 0.522 0.033 0.666 0.233 Peripherality - distance 1.288 0.002 1.489 0.020 Importance for higher-ranked centres 3.936 0.001 2.279 0.008 Suburban character (net migration) 1.406 0.259 0.776 0.316 Access to the nearest motorway 1.077 0.027 0.898 0.185 Functional polycentricity 1.052 0.843 0.809 0.336 Network density 0.982 0.959 2.537 0.001 N (cases) 113 166 b - 2LL 175.396 247.803 Nagelkerke R Square 0.529 0.479 Percentage Correctly Estimate 64.6% 64.5% Reference Category Low centrality a variables of historical importance and rural character are excluded because of their irrelevance in the case of BMA; b regression analysis for BMA includes all municipalities in order to ensure a sufficient sample size Note: Bold coefficients are statistically significant. Source: The South Moravian Region (2012), Czech Statistical Office (2013, 2016), MFČR (2013), Banky (2015), Czech Post (2015), CzechPoint (2015); author’s calculation

This model applies to centres with high centrality as well. In this case, service function is based even more on tourist attractiveness and the importance within the network of higher-ranked centres plays also a significant role. The model shows that the odd of having higher service function increases when a particular centre has a strong functional linkage with a higher-ranked

128 centre, i.e. small centre attract large number of commuters from a nearby large urban centre. Although the existence of in-coming flows seems to be a logical precondition for growth of the service function, the model points to a fact that especially the flows originated in large centres have a crucial impact on service function centrality of small centres. Considering the odds ratio value of the access to the nearest motorway is very close to 1, position towards motorway infrastructure does not have a considerable effect on service function centrality. According to the results of regression analysis for all small centres, it can be claimed that variables of historical importance do not have an impact on service function centrality. In other words, being a town or a centre supported by centrally-governed state during the socialist era does not necessarily mean that a particular centre still keeps also higher level of service function. Similarly, the variables indicating rural and suburban character of small centres are not associated with service function centrality when the entire region is analysed. Based on the results, interaction potential of small centres expressed by functional polycentricity and network density does not represent a significant variable. The service function is therefore not affected by the quality of public transport connections to large centres which is the factor entering the analysis as a potential precondition for borrowing size effect. Spatial pattern of the relative values of the index of service function centrality for small centres helps to imagine the results of regression analysis in a real territory (see Figure 4). The resulting spatiality of service function centrality together with the spatial characteristics of small centres suggests the existence of towns suffering from lying in the shadow of a larger centre (in terms of the own service functions) as well as towns benefiting from the proximity of higher-ranked centres and their workforce and borrowing size from larger urban centres. Small centres situated in the southwestern (e.g. Jevišovice, Mikulovice, Šumná) and the northern part (e.g. Lysice, Olešnice, Nedvědice) of the region as well as those located close to the BMA southern borders (Hustopeče, Klobouky u Brna), which is the area further away from large centres, represent municipalities whose high levels of service function centrality stem from their peripheral position where they serve as local centres. On the contrary, service function of small centres near Brno, Hodonín and Vyškov is lowered by the proximity of higher-ranked centres. Municipalities with the highest service function centrality are places with a strong tourist attractiveness. In this case, a location within an urban system is not so relevant. These are rather independent spatial units with a strong inner cultural and economic potential (e.g. Mikulov, Valtice, Lednice in the south; Vranov nad Dyjí in the southwest). Some of small centres situated in a close proximity to agglomerations of large urban centres are able to borrow size. A strategic location of centres Modřice and Rajhrad (south from Brno) or Chvaletice (south from Znojmo) is the main reason for still ongoing intensive commercial suburbanization taking place in these areas which contributes to higher levels of the service function. Regarding BMA, high service function centrality is closely related to the position of a particular municipality within the urban system network. If a municipality is sufficiently connected to other large centres (especially Brno) by public transport and attract commuters from the core city, service function centrality will more likely increase. Simultaneously, the service function tends to be higher when tourist potential increases and when a municipality is

129 situated further away from Brno. While certain municipalities profit from being a part of BMA, some centres in the Brno hinterland host less service facilities compared to other similar-sized municipalities within BMA. An agglomeration shadow of Brno affects municipalities that are located in the immediate neighbourhood of Brno and partially municipalities that have experienced a growth of population in recent years due to the process of residential suburbanization not supported by adequate increase of a number of service facilities (only when medium centrality is compared to low centrality; the case of e.g. Kuřim, Šlapanice, Mokrá- Horákov). However, the city of Brno provides a wide spectrum of urban functions substituting a lower service function of small centres to a great extent.

Fig. 4. Relative values of the index of service function centrality for small centres in the South Moravian Region. 1 – Předklášteří, 2 – Adamov, 3 – Boskovice, 4 – Černá Hora, 5 – Jedovnice, 6 – Křtiny, 7 – Kunštát, 8 – Letovice, 9 – Lomnice, 10 – Lysice, 11 – Olešnice, 12 – Ostrov u Macochy, 13 – Rájec-Jestřebí, 14 – Sloup, 15 – Svitávka, 16 – Šebetov, 17 – Velké Opatovice, 18 – Bílovice nad Svitavou, 19 – Dolní Kounice, 20 – Drásov, 21 – Hrušovany u Brna, 22 – Ivančice, 23 – Kuřim, 24 – Modřice, 25 – Mokrá – Horákov, 26 – Ořechov, 27 – Oslavany, 28 – Podolí, 29 – Pozořice, 30 – Rajhrad, 31 – Rosice, 32 – Říčany, 33 – Sokolnice, 34 – Střelice, 35 – Šlapanice, 36 – Telnice, 37 – Těšany, 38 – Tišnov, 39 – Újezd u Brna, 40 – Veverská Bítýška, 41 – Zastávka, 42 – Zbýšov, 43 – Židlochovice, 44 – Dolní Dunajovice, 45 – Drnholec, 46 – Hustopeče, 47 – Klobouky u Brna, 48 – Kobylí, 49 – Lanžhot, 50 – Lednice, 51 – Mikulov, 52 – Moravská Nová Ves, 53 – Podivín, 54 – Pohořelice, 55 – Rakvice, 56 – Sedlec, 57 – Šakvice, 58 – Valtice, 59 – Velké Bílovice, 60 – Velké Němčice, 61 – Velké Pavlovice, 62 – Vranovice, 63 – Archlebov, 64 – Blatnice pod Svatým Antonínkem, 65 – Bzenec, 66 – Čejč, 67 – Čejkovice, 68 – Dambořice, 69 – Dolní Bojanovice, 70 – Dubňany, 71 – Hovorany, 72 – Kyjov, 73 – Lipov, 74 – Lužice, 75 – Milotice, 76 – Moravský Písek, 77 – Mutěnice, 78 – Prušánky, 79 – Ratíškovice, 80 – Rohatec, 81 – Strážnice, 82 – Svatobořice-Mistřín, 83 – Šardice, 84 – Velká nad Veličkou, 85 – Veselí nad Moravou, 86 – Vracov, 87 – Ždánice, 88 – Bučovice, 89 – Drnovice, 90 – Habrovany, 91 – Ivanovice na Hané, 92 – Nesovice, 93 – Otnice, 94

130

– Pustiměř, 95 – Rousínov, 96 – Slavkov u Brna, 97 – Božice, 98 – Hrušovany nad Jevišovkou, 99 – Chvalovice, 100 – Jaroslavice, 101 – Jevišovice, 102 – Mikulovice, 103 – Miroslav, 104 – Moravský Krumlov, 105 – Olbramovice, 106 – Prosiměřice, 107 – Šatov, 108 – Šumná, 109 – Višňové, 110 – Vranov nad Dyjí, 111 – Dolní Loučky, 112 – Doubravník, 113 – Nedvědice Source: author’s processing

5. Concluding remarks The spatial dynamic of the South Moravian Region is strongly dependent on the regional capital Brno which can be considered a focus of rather monocentric pattern of the regional urban system. Although small towns (labelled as small centres in the paper) represent rather marginal part of the total number of municipalities, their role as centres for many smaller and rural settlements is therefore of even greater importance. A distribution of basic as well as higher- ranked services (as a number of facilities) has been analysed in order to evaluate the relationship between the service function of small centres and their position within a specific urban system. In line with the central place theory the services are spread in a territory according to their threshold and range reflecting their purpose and a hierarchical order of urban system. However, a large number of centres with similar population size differ in terms of their service function to a great extent. The study confirmed the assumptions arising from both borrowed size and agglomeration shadow concepts. A provision of services is affected not only by the population size of towns but it is also related to other factors explaining the differences of the service function of similar-sized towns. Service function centrality of small centres is associated predominantly with their importance for higher-ranked centres (interaction based on work commuting flows), tourist attractiveness (accommodation facilities) of places, and position of small centres towards large centres and population density of their hinterlands. Based on the results of the regression analysis the main findings could be summarized into following points: the service function is higher in centres situated in more peripheral areas where other stronger centre is mostly absent, and those situated in urbanized areas and in a close proximity to urban agglomerations are more likely to lack sufficient service function centrality (agglomeration shadow effect); the service function of small centres in the suburbs of BMA depends on the axis of the development of commercial zones, where small centres with higher service function centrality are located (borrowing size effect); being networked with large centres has a positive impact on service function centrality of small centres in BMA (borrowing size effect); the service function is higher in centres characterized by tourist attractiveness (fulfilling inner potential). The service function of rather peripheral centres with higher tourist potential is high despite these centres are not well embedded in regional network. On top of that, their centrality is not exposed to competition of other centres. In this context, the areas of a higher concentration of small centres are characterized by persistent hierarchical structure. The smallest municipalities provide only basic services and commuting to other larger centres is necessary. Regarding BMA, the most of towns in a close proximity to the main urban centre (Brno) do not borrow size in terms of the service function which is in accordance with findings by Meijers et al. (2016, p. 15) that ‘being well embedded in regional networks generally does not translate into a higher level of metropolitan functions’. Although some of these centres experience

131 population growth due to the process of residential suburbanization, their service function remains lower because of the competitive advantage of the nearest metropolitan centre. Using subdivision of borrowed size made by Meijers and Burger (2015), such centres ‘borrow performance’ (positive net migration in suburbs) but remain underserviced by selected functions. However, metropolitan functions can be observed in centres attracting commercial activities and work commuters from large urban centres. These centres are able to ‘borrow functions’ since they host more functions than other similar-sized municipalities. The service function deficit in centres affected by an agglomeration shadow is compensated by urban functions in a core city. A more frequent service-related commuting to the urban centre is an integral part of spatial-functional dynamic between the core and its hinterland. An evaluation of (dis)advantages stemming from such spatial configuration should reflect different levels of individual mobility as well as the access and ability to use technologies (transport, telecommunications, broadband access) that are crucial in production of socio- spatial relations. Centres with insufficient service function resulting from agglomeration shadow effect very often borrow performance (in terms of population growth). Hence, as Meijers and Burger (2015) point out, the debate about the concepts of borrowed size and agglomeration shadow should not lead to a dichotomous separation of both phenomena but rather to their interdependence and mutual conditionality. The results show that the complementarity of urban functions within the metropolitan area largely explains the relationship between the both concepts which verifies the assumption suggested by Burger et al. (2015, p. 1104) that ‘perhaps a place faces an agglomeration shadow in one respect, but borrows size in another’. While in the context of a metropolitan area a potential complementarity can be found between residential function and provision of higher-ranked services (specific cultural and social care facilities), rural areas with higher concentration of similar-sized small centres are characterized more by local specifics (spatial, physical, social) affecting mutual relations which probably mostly take form of an urban competition. However, an underlying mechanisms standing behind different forms of cooperation and competition need to be further investigated, preferably involving urban systems with preconditions for polycentric arrangement.

132

Paper 5: Questioning territorial cohesion: (Un)equal access to services of general interest

JIŘÍ MALÝ

Papers in Regional Science (2016), DOI: 10.1111/pirs.12250

ABSTRACT Although the debates about access to services of general interest have been at the core of the EU territorial cohesion discourse since the late 1990s, the impact of unequal accessibility to living conditions has yet be the subject of close inspection. The paper examines the relationship between the accessibility of services of general interest and demographic and socio-economic conditions in a specific Czech region. The analysis reveals inter-municipal disparities and identifies spatially excluded areas. Despite the negative association of insufficient access with education and depopulation characteristics, the results suggest that demographic and socio-economic development stems from a more complex set of factors.

Keywords: territorial cohesion, services of general interest, accessibility, spatial justice, Czech Republic

1 Introduction Territorial cohesion is a normative concept emerging within EU policies that has risen in importance, especially after the adoption of the Lisbon Treaty, when the territorial aspect of cohesion was added to its social and economic dimension (Doucet, 2006; EC, 2007b). In general, the territorial cohesion concept highlights the need to reflect on spatial inequalities within the EU regional policy, as well as better co-operation and strategies of sectoral policies with spatial impact. As a place-based approach, territorial cohesion accentuates the spatiality of social and economic issues contributing to the uneven pattern of spatial development (Davoudi, 2005), an integral part of more advanced forms of the societal development especially in capitalist economies (Hudson, 2015). From a normative political perspective, the territorial cohesion discourse contains several interpretations contributing to an uneasy explanation and comprehension of its real essence. The most recent debates are particularly characterized by a ‘cohesion’ and ‘competitiveness’ dichotomy (Evers, 2008; Vanolo, 2010). Despite plenty of more or less interrelated interpretations, territorial cohesion should be understood as a scale-dependent concept that is strongly influenced by specific socio-spatial contexts (ESPON, 2012b).

133

Although the concept of territorial cohesion has been widely discussed in the context of EU regional policy (e.g., EC, 2008; EU Ministers responsible for Spatial Development, 2011) and the academic sphere (e.g., Faludi, 2004; Evers, 2008; Servillo et al., 2012) in the last decade, most definitions struggle with vagueness and ambiguity. Next to the divergent ‘storylines’ framing the territorial cohesion discourse (Waterhout, 2007), the incomprehensibility of the concept is also caused by its multidimensionality (Doucet, 2006; Camagni, 2007). However, a core principle of the concept, profoundly embedded in territorial cohesion debates, includes the idea of fairness in terms of equal opportunities to all people regardless of their residence or job location (Molle, 2007). Thus, emphasizing fair access to jobs, education and basic services across the whole EU territory is one of the most significant parts of the territorial cohesion discourse. In the context of EU cohesion policy, the term services of general interest (SGI) has been established to label the basic services which should be provided by public authorities to ensure people’s essential needs (EC, 2003; EC, 2004b). The crucial importance of SGI is documented by their prominent position within the debates on territorial cohesion (see e.g. Camagni, 2009; ESPON, 2012b, 2013; Humer, 2014). Although SGI is not an official policy term, an appropriate combination of market and public services has ‘been at the core of the European integration process since its inception and remains the subject of a vigorous debate particularly in the EU-related legal and political science circles’ (ESPON, 2013, p. 8). According to the interpretation of territorial cohesion, which stresses the importance of balanced development, insufficient access to SGI is viewed as one of the most significant sources of spatial inequality. In this context, the improvement of socio-economic and demographic conditions in disadvantaged areas is closely linked to the level of SGI accessibility. However, transforming a broadly defined political aim of fair access to SGI into a more analytical level of spatial planning practises seems to be an uneasy process. Although the role of accessibility in spatial development is deep-seated in academic literature and in the overall debates about European spatial development (e.g., Keeble et al., 1982; Vickerman et al., 1999; López et al., 2008), the nature of the relationship between accessibility and socio- spatial disparities has not been a subject of wider discussion regarding territorial cohesion since its inception. Reflecting the importance of fair access to SGI, this paper aims to clarify the socio- spatial interdependencies framing spatial differences in access to socially valued resources (labelled as SGI within the terminology of EU regional policy) at the regional level. The paper argues that unequal access to SGI is an aspect that plays a significant role in forming socio- spatial inequalities. Therefore, the study focuses on questions of whether and how the level of SGI access is related to demographic and socio-economic conditions of particular places (municipalities) and how this relationship can be interpreted with respect to the structure of the urban system and functioning of the transport infrastructure. The South Moravian Region in the Czech Republic serves as a case study to investigate this relationship.

134

2 Theoretical background 2.1 Cohesion, equality and justice The idea of balanced spatial development is deeply rooted in the debates about regional and spatial planning at the EU level (Kunzmann, 2006; Faludi, 2010). So much so that the concept of territorial cohesion appeared simultaneously with works on drafting the European Spatial Development Perspective (EC, 1999). With regard to the need to create a long-term strategy that would make the spatial development of the European territory more efficient, territorial cohesion seemed to be the proper concept to ensure balanced spatial development. Besides other interpretations of territorial cohesion, the fundamental storyline ‘Europe in Balance’ (as identified by Waterhout, 2007), which combines the French spatial planning tradition ‘aménagement du territoire’ (regional economic approach) with an integrated comprehensive approach (applied in Germany), is focused on levelling regional disparities (Faludi, 2004). From this perspective, a similarity with socialist spatial planning tradition can be observed (Malý and Mulíček, 2016). The Territorial Agenda of the EU 2020 (EU Ministers responsible for Spatial Development, 2011, p. 2) states that territorial cohesion ‘enables equal opportunities for citizens and enterprises, wherever they are located, to make the most of their territorial potentials’. However, the meaning of ‘equal’ is not questioned any further and remains ambiguous. According to Davoudi (2005), such an interpretation of territorial cohesion added a spatial justice dimension to discussions about spatial planning at the EU level. Lévy (2003), cited in Peyrony (2007) argues that spatial justice is reflected within the territorial cohesion discourse by extending and applying the Rawlsian principle of equity to territorial development. Specifically, the comprehensive Rawlsian liberal democratic ‘theory of justice’ is based on the principle of ‘liberty’ (everyone has the right to the same basic freedoms); the principle of ‘fair equality of opportunity’ (social and economic inequalities are not acceptable unless they are associated with offices and positions that are open to everyone); and the ‘difference’ principle, which accepts inequalities should they most benefit those who are the least advantaged (Rawls, 1971). However, from a geographic point of view, the universal and normative character of the theory of justice does not reflect the distribution of inequalities through space, the specifics of different societies and cultures, nor the temporal aspect of development in modifying the level of spatial disparities. A more radical critique then emphasized only the static view of social inequality embedded in the theory of justice, ignoring the causes of inequalities and the structural processes lying behind them (Soja, 2010). From a spatial planning perspective, the spatialization of the fairness principle, an integral part of Rawlsian theory, is expressed through the notion that people should not be disadvantaged by their residence or work location (EC, 2004a; Molle, 2007). In this context, the concept of territorial cohesion implies that the place of residence and work affects living standards and quality of life in itself. The crucial issue concerning the rights of individuals is the preservation of public goods where a market-based provision fails (ESPON, 2013). Ensuring adequate access to basic services is accentuated primarily by the Territorial Agenda of the EU 2020 (EU Ministers responsible for Spatial Development, 2011, p. 8) declaring that ‘fair and affordable accessibility to services of general interest, information, knowledge and

135 mobility are essential for territorial cohesion’. The need for fair access to services originates from a simple awareness of spatially expressed inequalities among municipalities (or other territorial units) regarding basic facilities and public services (e.g., Davoudi, 2005; Molle, 2007). As Soja (2010) argues, despite the efforts to equally distribute socially valued resources across a territory, it is practically impossible to achieve equal access to these resources. Therefore, a certain level of distributional inequality is inevitable (Soja, 2010). If other factors are added to basic distributional inequality (e.g., institutional inefficiency, budgetary requirements, personal greed, racial intolerance or abuse of power), then inequality is further accentuated. A combination of several factors may, after exceeding certain acceptable limits, cause a feeling of injustice and human rights violations. Even when the ideas about homogenization of space (in terms of equal access to services) are replaced by a more realistic vision of reaching adequate and fair access in disadvantaged regions, the acceptable levels of inequality are still questioned (Handy, 1992; Witten et al., 2003; ESPON, 2013). Following the territorial cohesion logic that stems from the Rawlsian philosophical macro-theory and realizing that differences in access to basic services are (and will be) an integral part of the spatially heterogeneous European territory, undesired spatial disparities in terms of the accessibility occur only when the socio-economic conditions of a particular place decrease as a result of inadequate access to socially valued resources. Hence, territorial cohesion is not achieved so long as insufficient access to services continues to deteriorate the living conditions in an affected territory. Besides the spatial dimension of inequality, the promotion of fair access to resources is an indirect consequence of the neoliberal principles applied in the EU. The normative narratives in the territorial cohesion discourse are embedded in this economic model, which is still predominant in EU regional policy. Thus, reflecting the fundamental logic of competitiveness and economic growth in capitalist economies and using the perspective of theories focusing on uneven development regardless of their theoretical background (e.g., Hirschman, 1958; Harvey, 1982; Krugman, 1991), the principle of social solidarity might be considered as a way to mitigate the socio-spatial inequalities caused by the growing neoliberalization tendencies within the territory of the EU (Hudson, 2015). According to Farrugia and Galina (2008), the political objectives of economic growth and social cohesion are not even compatible within the framework of the dominant economic model. Similarly, Davoudi (2007) perceives the political decoupling of European economic integration and social-protection issues as an inherent and persistent tension between the EU economic competitiveness and cohesion policies. From such a perspective, promoting equal access to SGI is a result of the gradual integration of social and just principles into spatial development policies. In terms of the theories of justice, such tension reminds us of the contradiction between the proponents of the Rawlsian theory of justice and the libertarian-minded critics stressing the need for a free market which ensures personal property without limit, and who also criticize the excessive redistributive mechanisms restricting individual rights and freedoms (Nozick, 1974).

136

2.2 Access to services of general interest Within the debates concerning spatial planning at the EU level, the concept of spatial justice has been operationalized through the notion of fair access to SGI. According to the Green Paper on Services of General Interest (EC, 2003, p. 3), SGI ‘touch on the central question of the role public authorities play in a market economy, in ensuring, on the one hand, the smooth functioning of the market and compliance with the rules of the game by all actors and, on the other hand, safeguarding the general interest, in particular the satisfaction of citizens’ essential needs and the preservation of public goods where the market fails’. SGI concerns a wide range of activities, both economic (market) and non-economic (non-market) in nature, that are usually provided in the public interest by the state, regions or municipalities (CEEP 2010). However, more precise characteristics of SGI are lacking (see e.g. Rauhut and Ludlow, 2013, who provide an overview). The concept of SGI is derived from the more specific term services of general economic interest (SGEI), which is focused on market services. Almunia (2011), cited in Ølykke and Møllgaard (2013, p. 206) argues that ‘services of general economic interest are activities that would not be produced by market forces alone or at least not in a form that would be available to all’. In general, SGI is a broader concept that includes services provided by governmental or public administration bodies (e.g., education systems, social protection, healthcare) and services that are partially provided by the private sector and are profitable, that is, SGEI (e.g., cultural facilities, postal services, waste management, telecommunications or transport infrastructure). The study of the ESPON programme adds social services of general interest (SSGI) to SGEI, including labour market services, education, healthcare, childcare, elderly care, social care, (social) housing and social assistance services. SGI are then defined as SGEI and SSGI (ESPON, 2013). Such an ambiguous definition of SGI/SGEI is due to the remaining responsibility of the EU member states to organize and provide these services (EC, 2003). Each EU member state (or region) can define SGI with regard to the specific legal framework and socio-spatial arrangement of the territory. Adequate access to SGI cannot be achieved without the proper functioning of transport infrastructure, according to the European Spatial Development Perspective (ESDP), which emphasizes parity of access to infrastructure and knowledge (EC, 1999). It has been argued that policies ‘should also ensure that high quality infrastructure, for instance high-speed/high- capacity rail lines and motorways, do not lead to the removal of resources from structurally weaker and peripheral regions (‘pump effect’); or that these areas are not crossed without being connected (‘tunnel effect’)’ (EC, 1999, p. 26). The basis for better accessibility, overall balanced spatial development, and a strengthening of the of the peripheral and less favoured regions’ competitiveness should be polycentric development1 (EC, 1999; EU Ministers responsible for Spatial Development, 2011), despite the missing empirical evidence of its positive impact (Meijers and Sandberg, 2006; Malý, 2016).

1 Thus, polycentric urban systems are supposed to enhance the internal regional development by combining cohesion (access to SGI) and competitiveness (integrated economic zones) goals emerging from the territorial cohesion discourse.

137

The routing of transport and telecommunication infrastructure together with the structure of the urban system are key factors contributing to the accessibility of SGI. With regards to urban system theories, the original ‘central place’ models (Christaller, 1933; Lösch, 1944) were first transformed into more complex theories respecting the specifics of particular urban systems (Berry, 1964), then into ‘network’ models (Camagni, 1993) and ‘polycentric structures’ (Kloosterman and Lambregts, 2001; Parr, 2004), which are connected to EU spatial planning documents. The urban networks have evolved into multifaceted systems strongly influenced by the globalization process, continuing metropolization, and the enlargement of the membership in the case of the EU. In spite of the undisputed role of central place theory in terms of providing a framework for identifying the threshold and the range of particular services and goods, the spatial disparities of access to SGI might be the result of more complex factors including, for example, functional dynamics within an urban system, agglomeration effects (see Partridge et al., 2009 for the ‘agglomeration shadow’ discussion and Meijers et al., 2016 for the ‘borrowing size’ issue), and historical development (e.g., transformation of the economy in post-socialist countries). With the growing importance of the largest agglomerations in national (regional) economies, the traditional hierarchical structure of urban systems has experienced a shift towards the formation of complex networks of spatio-functional interdependencies (Camagni, 1993). Hence, the position of particular places within urban system networks, and especially their ability to benefit from the advantages provided by metropolitan areas and urban agglomerations, is essential for their development. In this context, a locality lying in an agglomeration shadow may suffer from the proximity of a higher-ranked centre (e.g., in terms of providing insufficient basic services) or, in some cases, a particular place could ‘borrow’ functions from its larger neighbour (Burger et al., 2015). At the regional level, the impact of transport infrastructure on the functions of particular places is undisputable. Thus, the transport infrastructure and the functioning of the urban system are both aspects reflected in the following sections aimed at clarifying the relationship between access to SGI and living conditions of particular places.

3 Methodology and data 3.1 Research objectives and study area The main research objective of the paper is to reveal the relationship between the access to SGI and the level of socio-economic disadvantage. While spatial differences in access to SGI are expected, and do not serve as a proof of an incohesive situation, the association with the level of socio-economic conditions of particular places provides a deeper understanding of the territorial cohesion issue. The assumption of an association between improved access to SGI and the socio-economic level of development is evaluated through a quantitative analysis based on a robust database that uses secondary data. There are two main steps in the analysis. First, SGI accessibility reflected in car time distance, public transport connections, and public transport fare is calculated. Second, SGI accessibility for each type of distance is correlated with the socio-economic levels of particular places and then subjected to a multiple regression analysis. The regression analysis assesses whether SGI accessibility (predictor variables) accounts for the variability in a socio-economic disadvantage (dependent variable). The South

138

Moravian Region (Czech Republic) has been chosen as a case study due to the diverse nature of its urban structure and the occurrence of both central and peripheral locations in terms of economic performance.

Fig. 1. The South Moravian Region. Source: Czech Statistical Office (2014c); author’s processing

The South Moravian Region is situated in the south-eastern part of the Czech Republic (see Figure 1). Municipalities within its official administrative delimitation have been set up as spatial units in the analysis. The South Moravian Region belongs to one of the largest regions in the Czech Republic with the number of its inhabitants reaching almost 1,200,000 and a total area of over 7,000 km2 (Czech Statistical Office, 2014c). The character of the urban system is especially determined by its regional capital Brno (where approximately one-third of the regional population lives), playing a key role in the economic, administrative and educational development of the entire region (Mulíček and Toušek, 2004). Additionally, the Brno metropolitan area is characterized by strong transport integration based on transportation flows (Seidenglanz, 2010). The monocentric urban structure is also fostered by the absence of a similar-sized centre beyond the regional borders. Cross-border relations are strongly limited, particularly in the southern direction, where the national border still represents a barrier for most of the Czech people. While the Austrian border still acts as a linguistic barrier to some extent, relations with Slovakia are rather one-sided, with a prevailing direction of flows from Slovakia into the Czech Republic (in terms of commuting patterns and employment characteristics).

139

After the fall of the iron curtain in 1989, the Czech Republic experienced a significant social and economic transformation which has also been related to changes in the urban system functioning. In the South Moravian Region, the changes in migration and employment patterns influenced commuting forms and the centrality of municipalities. The socialist period was characterized by a deep differentiation in the population growth of municipalities. The collectivization of agriculture and massive industrialization led to the growth of municipalities, especially in size categories from 5,000 to 20,000 inhabitants, and to the decline of smaller, non-central municipalities (Hampl, 2005). However, the democratization of the society has brought new growth to the Brno metropolitan area and a decrease in the importance of regional secondary centres. The region has been affected by processes of residential and commercial suburbanization (like other regions in post-socialist space, see Kovács and Tosics, 2014; Kurek et al., 2015 or Nuga et al., 2015), which has especially occurred in the hinterland of Brno (Mulíček, 2002). On the other hand, stagnation and sometimes loss, in terms of migration rate, are typical characteristics of other larger city centres within the region. The transformation of the society’s organization has had an impact on the service functions of settlements and consequently on access to basic services. Based on the transformation processes, SGI could be placed unfairly and with a deeper impact on territorial cohesion, particularly for peripheral localities. The peripheries of the South Moravian Region are represented by rural areas where the local centres are losing their position in the service processes (mostly shopping function, see Maryáš et al., 2014 or Kunc et al., 2016); however, such development is typical for other countries as well (e.g., Cebollada, 2009; Wang and Tang, 2013).

3.2 Calculating the accessibility of SGI In the Czech Republic, the term SGEI should be understood as economic activities that public authorities define as services of special importance to its citizens and that are not provided without public intervention; their provision reflects a market failure (MMR, 2013). The official definition of SGEI is very brief, stating that SGEI is a ‘service which is provided universally and permanently and the special legal act grants the provider special rights and obligations’ (Parliament of the Czech Republic, 2009). Consequently, SGI is the overarching term including SGEI and services provided outside of the market and on a free-of-charge basis, for example, compulsory education and social systems, health and culture non-commercial services (Halásek, 2012). From the national perspective, SGI are described very broadly. The choice of particular services, which could be considered SGI in this analysis, is thus predominantly influenced by the potential importance of services in daily human lives, by their public support2 and by the availability of data (source of data: The South Moravian Region, 20123; Czech Statistical Office, 2013; websites of banks and insurance companies operating in the Czech Republic – access through Banky, 2015; Czech Post, 2015; CzechPoint, 2015). The final list of

2 In some municipalities, the service is operated only by the private sector on a commercial basis (e.g., cinemas). However, if there is at least one municipality that supports the provision of this service by public tools, the service is categorized as SGI. 3 The database of facilities stemming from the survey made by the South Moravian Region (2012) is not complete. The missing data was gathered by exploring municipality websites, field research and local expertise. Thus, the final database consists of data from several sources which might cause a slight inaccuracy in the measurements. However, this inconsistency has no effect on the results due to the scale of the analysis.

140

SGI covers the most fundamental human needs. SGI include health care facilities, social services, education facilities, retail services, special services and cultural amenities. Each of the typological groups includes five services4. The presence of services has been detected in each of the 672 municipalities in the South Moravian Region. Additionally, the quantitative significance of each service (measured by its occurrence in the municipality, i.e., the number of facilities) has been identified. The sum of the number of facilities F has been calculated for each service. Based on a mutual comparison of the quantitative significance (the number of facilities F) among all services, the score of the centrality SC, as an indicator of the qualitative significance of the service, has been determined. The SC with the value 1 has been assigned to the most common service in terms of the number of facilities in the entire region, marked as F1. Thus, the service with the highest number of facilities represents the lowest level of SC and the service characterized by the lowest occurrence gains its qualitative significance by a higher SC. In other words, the most spread- out services with a lower SC create small catchment areas; they are used on a rather daily basis, and are typical of less important centres (e.g., grocery shops, pubs, basic education). However, a higher SC indicates specialized services mostly concentrated in major centres and used less frequently (e.g., theatres, financial institutions, special health care facilities). We can imagine a model situation when SC of three services (a, b, c) is calculated in a territory composed of municipalities i and j (see Table 1) and where the number of facilities F for each service in the whole territory is known. The service with the highest F, in this case service a, is evaluated by SC 1 and as the F1 value enters into the following calculation of SC for the other service b:

= 1 𝑏𝑏 𝐹𝐹 𝑆𝑆𝑆𝑆 𝑏𝑏 If the values from Table 1 are added to 𝑆𝑆 the equation, SC for other services can be calculated. Based on SC values and known F for municipality i the municipal centrality MC for service a and municipality i merge into: = ( )

𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎 where Fai means the number of facilities𝑀𝑀𝑀𝑀 F for 𝐹𝐹the particular𝑆𝑆𝑆𝑆 service a and municipality i, SCa is the score of centrality SC for the particular service a. The municipal centrality MCa is a dimensionless number consisting of the presence and quantitative significance of the particular service a in each municipality. The higher the value of MC, the higher the centrality.

4 Health care facilities – general practitioner, paediatrician, dentist, hospital, emergency medical service; social services – retirement home, facility for physically disabled people, shelter, low-threshold facility for children and youth, nursing home; education facilities – early childhood education, primary education, low secondary education, upper secondary education (grammar school as a separate category); retail services – grocery shop, drugstore, shopping centre, pub, restaurant; special services – post office, bank, ATM, insurance company, CzechPoint (assisted place of public administration); cultural amenities – public libraries, cultural centres, cinema halls, museums and galleries, theatres.

141

Tab. 1. The model values of the number of facilities F, the score of centrality SC and the municipal centrality MC for services a, b, c and municipalities i and j. Municipalities Services F (sum) SC (sum) F MC i j i j

a 30 (F1) 1 25 5 25 5 b 20 1.5 10 10 15 15 c 10 3 2 8 6 24 Source: author’s calculation

Accessibility is understood in its broadest sense, that is, the ability of a particular transport system to reach a place of desired activity from a current location (Dalvi and Martin, 1976) which relates to a potential spatial exclusion (Preston and Rajé, 2007). Overcoming physical distance through modes of transportation is therefore a crucial aspect in the evaluation of accessibility. SGI accessibility has been measured using time distance by car, public transport connections and public transport fares (see Table 2 for a detailed description of the methodology). Distance d has primarily been measured with the three nearest centres of a particular service for two reasons. First, the connection to only the first nearest centre would not include the plurality of choices or a broader spatial context. Second, including connections to all centres in the country would distort the spatial framework of daily human lives organized in rather contained local spatial arenas. Intra-national cross-border relations have been taken into account (trans-national relations have not been considered due to the limitation of the data availability). If the municipality provides a specific service, it is included in the centres. Access to the service a has been determined by using the distance d to the three nearest centres j, k, l (ordered from j as the nearest) and the values of MC in these three municipalities for a particular service a (the calculation is inspired by Maier et al., 2010). In the case of car time distance and public transport fare, the accessibility T of a particular service a and a municipality i gives the equation5: = / + / + / + / /4

𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 In the case𝑇𝑇 of public𝑀𝑀𝑀𝑀 transport𝑑𝑑 𝑀𝑀 connections,𝑀𝑀 𝑑𝑑 𝑀𝑀 the𝑀𝑀 accessibility𝑑𝑑 𝑀𝑀𝑀𝑀 T� of𝑑𝑑 a particular� service a and a municipality i gives the equation: = + ( ) + ( ) + 4

𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎 Apart from𝑇𝑇 including𝑀𝑀𝑀𝑀 �𝑑𝑑 connections� 𝑀𝑀𝑀𝑀 𝑑𝑑 to the 𝑀𝑀 three𝑀𝑀 𝑑𝑑 nearest 𝑀𝑀municipalities𝑀𝑀 � 𝑑𝑑 � (j, k, l) the accessibility T also considers the own value of MC for municipality i. The purpose is to reflect the presence of a particular service in the evaluated municipality which has an impact on the overall accessibility level, that is, if the service is available in the municipality for which the

5 As the public transport connection differs from time distance and public transport fare by informative meaning (i.e., more public transport connections means higher accessibility but larger distances and higher fares mean lower accessibility), the calculation of T had to be modified in order to make the T value clear. Subsequently, the higher the T value, the better the access in terms of public transport connection (a higher number of connections), car time distance (shorter distance) and public transport fare (cheaper tickets).

142 level of access is calculated, it is reflected in the equation. Such a municipality is then favoured at the expense of the municipality where people are dependent on commuting to other centres. In fact, the accessibility T indicates the potential for localization, enabling interaction with other centres. The T value basically reflects the attractiveness of places (municipalities) in terms of the presence of and access to a specific service, as determined by the spatial configuration of the nearest centres and infrastructure. The higher the accessibility T, the better the access to the particular service. The summarization of all T values for one municipality i merges into an overall SGI accessibility indicator TSGI:

= 𝑛𝑛

𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 � 𝑇𝑇𝑎𝑎𝑎𝑎 𝑎𝑎=1 Tab. 2. The methodology used for calculating each type of distance d between the municipality and the centre of a particular service. Type of distance d Calculation of distance d

. The fastest connection between the centres of municipalities in Time distance by car minutes (GIS software calculation) . Calculation respects speed limits, road categories and terrain

. The number of public transport connections (buses, trains and city public transport connections) linking the municipality to the centre of a particular service during a weekday (to the date: 18 March 2015; IDOS, 2015) Public transport connections . The connection includes one transfer at most; the advantage of a direct connection is supported by the calculation – the transfer connection is by one-third smaller (instead of counting it as 1, its value is 0.33)

. An average ticket price of public transport connecting the Public transport fare municipality with the centre of a particular service based on tariff zone prices (IDS JMK, 2015) Source: author’s processing

3.3 The relationship between SGI accessibility and the level of development The next step of the analysis aims at revealing the association between SGI access and demographic and socio-economic conditions of particular municipalities. For this purpose, the indicators characterizing these conditions have been determined. The demographic situation is described by population ageing characteristics such as age index and old-age dependency ratio, the attractiveness of particular municipalities by population growth, the crude net migration rate and completed dwellings (calculated between 2001 and 2011). The socio-economic conditions are represented by the share of university-educated people in the total population over 15 years, employment in the tertiary sector and the unemployment rate. The source of statistical data is the Czech Statistical Office (2014a). All indicators have been standardized by the Z-score

143 method and transformed into a comparable form (the Z-score of 0 has been given the value of 100 and the standard deviation of 1 equals the value of 20, for example, 1 = 120, 0.5 = 90). Accordingly, as the value of the standardized indicator increases, the particular characteristic becomes more desirable. The Pearson correlation coefficient has been used to evaluate the relationship concerning SGI accessibility and partial aspects of development. In order to reduce the scope of the data, the levels of access to the particular services have been grouped into three categories based on the SC values of the SGI (the sum of the values). The potential bias in the results caused by outliers has been eliminated by a log-transformation of all variables. The SGI accessibility scores serve as the explanatory (independent) variables that are included in the multiple regression analysis. The expectation is that the demographic conditions, the attractiveness of the places in terms of population growth and new construction, and socio-economic conditions as well (these are the dependent variables in the regression analysis) change when access to SGI varies. The aim is not to reveal causality (which is not even possible using such an analysis) but rather to provide statistical evidence about the form of dependency between the two variables. Another reason for grouping SGI accessibility into three categories is due to the multicollinearity that appears when the accessibility of each service enters the multiple regression analysis. The categorization of SGI accessibility according to SC values prevents this multicollinearity effect. In meeting the preconditions of linearity and homoscedasticity the stepwise multiple regression analysis has been conducted. The following part describes the main results of the analysis.

4 Results 4.1 The score of the centrality SC, the municipal centrality MC and the accessibility T The quantitative significance of a particular SGI is strongly dependent on its purpose and potential utilization. The most common services fulfilling basic human needs are characterized by the lowest SC because of their intensive spatial distribution and density (see Table 3). Quantitatively, the most well-distributed SGI (higher S value) favour daily use (e.g., pubs, restaurants, grocery shops), which indicates a frequent demand for these services and thus smaller catchment areas as well as limitations to their influence. In contrast, social services and health care facilities represent SGI with a rather sparse spatial distribution. The function, the importance for human lives and living conditions, the temporal aspect of usage and the overall nature of SGI situated at the opposite ends of SC hierarchical order differ to a great extent. An unbalanced distribution of a particular SGI is in accordance with the central place theory, at least when its general aspects are taken into account. The municipal centrality MC stems predominantly from the hierarchical structure of the urban system (an almost perfect correlation between MC and the population size of municipalities). However, the population size of municipalities should not be perceived as the only factor affecting the level of MC. In fact, the spatial localization within a specific urban system may have a strong impact on the provision of SGI in a particular municipality.

144

Tab. 3. The score of centrality SC for a particular SGI; municipalities in the South Moravian Region. SGI typological SC SGI F SC group a category Pub RS 3,213 1.0 3 Restaurant RS 1,689 1.9 3 Grocery shop RS 1,603 2.0 3 CzechPoint SpS 807 4.0 3 Public library CA 751 4.3 3 Dentist HF 694 4.6 3 Early childhood education EF 644 5.0 3 Bank SpS 637 5.0 3 General practitioner HF 621 5.2 3 Cultural centre CA 606 5.3 3 Primary education EF 471 6.8 3 Insurance company SpS 432 7.4 3 ATM SpS 431 7.5 3 Post office SpS 361 8.9 3 Paediatrician HF 309 10.4 3 Drugstore RS 276 11.6 3 Low secondary education EF 272 11.8 3 Upper secondary education EF 180 17.9 2 Museum and gallery CA 178 18.1 2 Nursing home SS 105 30.6 2 Cinema hall CA 87 36.9 2 Retirement home SS 47 68.4 2 Grammar school EF 41 78.4 2 Theatre CA 30 107.1 1 Shelter SS 23 139.7 1 Emergency medical service HF 22 146.0 1 Hospital HF 21 153.0 1 Facility for physically disabled people SS 18 178.5 1 Low-threshold facility for children and youth SS 18 178.5 1 Shopping centre RS 14 229.5 1 a RS = retail services; SpS = special services; CA = cultural amenities; HF = health care facilities; EF = education facilities; SS = social services Source: The South Moravian Region (2012), Czech Statistical Office (2013), Banky (2015), Czech Post (2015), CzechPoint (2015); author’s calculation

Medium-sized or smaller towns near the regional capitals may ‘suffer’ from close proximity to the higher ranked centres. In some cases, intense commuting and consumption of services in the regional centre have a devastating effect on the local service providers. The competitive advantage of the regional capital becomes the reason for the insufficient provision of SGI in these towns. In the case of the South Moravian Region, based on the levels of MC, this pattern can be observed in small towns situated within the Brno metropolitan area (i.e., Kuřim, Šlapanice, Rosice). On the other hand, these municipalities benefit from their proximity to Brno with respect to the accessibility of SGI with the highest scores of the centrality SC. The levels

145 of MC also show that even smaller towns act as local centres when they are situated in more peripheral areas (e.g., Klobouky u Brna, south-east of Brno, or Velká nad Veličkou in the south- eastern border area), and peripheral municipalities can serve as centres for higher ranked SGI, especially when social services are taken into account (the reason could be the localization of retirement and nursing homes into more peaceful and cleaner environments). The accessibility T calculated as the access to the three nearest centres by public transport, varies depending on the position of the municipality within the public transport network, and logically on the particular SGI that is being evaluated. Nevertheless, the summarization of the particular SGI provides a picture illustrating overall SGI accessibility TSGI with a spatial pattern of access to SGI. The central position of Brno is demonstrated by the highest value of TSGI (see Figure 2). Municipalities located in the Brno hinterland benefit from locations that are intersected by direct connections from Brno to many other local centres in the region. Therefore, the municipalities in close proximity to Brno are characterized by higher values of TSGI. The data suggests that municipalities connected to major transport routes are favoured simply due to their spatial localization. As can be seen from the map output (Figure 2), the major transport routes attract a higher number of public transport lines (long-distance transport lines leading to Brno play a significant role), which results in the increasing number of public transport connections in the municipalities that are part of such infrastructure. The originally separate bus lines operating in the peripheral areas often meet and intersect major roads. However, this does not apply to municipalities excluded from transport routes by missing slip roads (as is the case of the motorway connecting Brno to the southern part of the region). In peripheral areas, railway proximity is more significant. Bus service is an insufficient alternative to train connections in municipalities which do not benefit from railway infrastructure. Such a disadvantage is striking especially in the northern area of the region, where the municipalities situated further away from the railway corridor are spatially excluded from access to SGI. Likewise, in the south-eastern part of the region, the municipalities (small rural villages) situated close to the national border do not benefit from the not-so-distant railway corridor at all. The third large area of spatially excluded municipalities is situated on the southwestern regional border, where a higher number of more important centres is absent.

TSGI has also been calculated using public transport fares. Although the general spatial pattern of the TSGI levels does not differ from the same measurement based on the public transport connections to a great extent, some partial discrepancies can be found. Regarding the Brno metropolitan area (Brno and the nearest municipalities), the financial cost of reaching the nearest SGI centres is slightly higher in the Brno hinterland. Despite the lower quality of public transport connections, these small municipalities have a similar transport fare as the regional capital. With increased distance from the central part of the region, TSGI (in terms of public transport fares) decreases. In contrast to public transport connections that are not primarily based on mutual distance, the public transport fare is higher in localities situated further away from the SGI centres. While even small municipalities in the central part of the region suffer from indirect connections to SGI centres (Brno often serves as a place of transfer) but their proximity to SGI centres has an impact on lower fares, more distant areas are disadvantaged both by indirect connections and higher prices of transport due to the greater distances from the nearest SGI centres.

146

When focusing on car time distance, the higher number of well-connected municipalities within the Brno metropolitan area becomes quite evident. The dependence of automobile transport on road infrastructure is essential. Thus, even municipalities excluded from efficient public transport connections can benefit from proximity to highways and major roads. The advantage of car users to overcome greater distances is also evident in border areas where the connection to major roads allows the municipalities to reach the SGI centres situated outside the regional borders. On the other hand, as opposed to the public transport connection and fare, an inner ‘peripheral pocket’ can be observed southeast of Brno due to the lack of backbone roads.

Fig. 2. The SGI accessibility TSGI; municipalities in the South Moravian Region. Source: IDOS (2015), IDS JMK (2015); author’s calculation and processing

4.2 SGI accessibility and its relation to the level of development The uneven spatial distribution of SGI together with distinctive opportunities to reach services, the result of unequal connections to the public transport network and transport infrastructure, are supposed to be the factors which contribute to a less balanced spatial development. However, within the framework of the territorial cohesion discourse, the impact of SGI accessibility on the socio-economic condition of specific localities has not yet been evaluated. Based on the analysis of SGI accessibility we can identify localities that are more spatially excluded from access to the nearest SGI centres. The question is whether these localities are also characterized by more unfavourable social, economic or demographic conditions. In order to examine such a relationship, the Pearson correlation coefficient between the values of

147 accessibility T for SC categories and the indicators of demographic and socio-economic conditions were calculated (see Table 4). With respect to the standardization of indicators, the correlation of variables should be more easily interpreted. In other words, the higher the standardized indicator, the more favourable the value of the particular characteristic (i.e., in the case of the unemployment rate and age index, higher standardized values indicate lower levels of original values). Consequently, the positive correlation means that the demographic or socio- economic conditions grow with increasing TSGI. As is common with large samples, the results show a statistically significant correlation in most cases. However, the correlated values indicate a rather low positive relationship between levels of accessibility for all types of distances and population ageing characteristics. Therefore, we can assume that spatially expressed differences of the ageing process do not relate to SGI accessibility. Regarding the attractiveness of places and socio-economic conditions, the positive and mostly mild correlation shows that there is a relationship (although not so strong) between improved access to SGI and higher levels of measured development indicators. Only in the case of access to the services of SC 2 category, using car time distance in the calculation, can we observe a substantially strong correlation with the presence of a university-educated population. A closer examination of the mutual relations of variables is provided by the multiple regression analysis. A low to mild correlation of variables suggests that the association of SGI access with the level of development should not be overestimated. The multiple regression analysis is used to evaluate the level of association of the explanatory variables with the attractiveness of places and socio-economic conditions. The indicators of population ageing are not included since their correlation with the explanatory variables has not been proved (at least in a linear form, which is essential to the presented analysis). As the results of the multiple regression analysis for each type of distance, and thus accessibility (public transport connections, fare and car time distance), do not differ to a great extent (similar to the correlation values), the empirical findings will be described using the data on car time distance SGI accessibility. For each predictor, the model explaining the highest percentage of the explained variable is followed by a model operating with the strongest explanatory variable (see Table 5). Besides the spatial context, the temporal aspect is reflected by the driving time needed to cover the distances between municipalities, which is mainly based on the quality of transport infrastructure (i.e., motorways, major roads). Based on the three different levels of access (SC 1, SC 2 and SC 3), characterizing the hierarchy of SGI in terms of the size of their catchment areas, we can examine their contribution to the level of development. While the attractiveness of places expressed through population change is predominantly influenced by access to services with a higher range of serviceability (in comparison to other categories), the socio- economic conditions are characterized by more variation regarding the influence of particular SC categories. However, according to the results of the multiple regression analysis, the explanatory power of the predictors (accessibility values T for all SC) is relatively low (as the values of the adjusted R2 indicate). Thus, the variability of the development level can be explained by access to SGI only partially, that is, there are other factors contributing more to the attractiveness of places and their socio-economic conditions.

148

Tab. 4. The Pearson correlation coefficient between the accessibility of services based on SC category (SC 1 = 100<; SC 2 = 15-99; SC 3 = 0-14) and the development indicators; municipalities in the South Moravian Region. T Indicators of development (standardized characteristics) Population ageing Attractiveness of places Socio-economic conditions Age Old-age Population Crude net Completed University- Employment in Unemployment

index a dependency ratio b growth c migration rate c dwellings d educated e tertiary sector f rate

Public transport connection SC 1 .147** .148** .324** .229** .294** .473** .452** .307** SC 2 .133** .139** .338** .254** .315** .482** .445** .300** SC 3 .135** .139** .336** .255** .328** .498** .441** .250** Public transport fare SC 1 .102** .103** .452** .374** .359** .491** .307** .407** SC 2 .077* .090* .417** .368** .364** .471** .328** .333** SC 3 .108** .117** .380** .323** .365** .490** .374** .219** Car time distance SC 1 .102** .105** .481** .430** .384** .484** .374** .421** SC 2 .101** .075 .401** .361** .353** .549** .426** .298** SC 3 .181** .165** .210** .105** .212** .451** .470** .217** * correlation is significant at 0.05 level; ** correlation is significant at 0.01 level; a the number of persons 65 years and over per one hundred persons under age 15; b the number of persons 65 years and over per one hundred persons 15 to 64 years; c for the time period 2001-2011; d the number of completed dwellings per one thousand inhabitants for the time period 2001-2011; e the share of university-educated inhabitants in the total municipal population over 15 years; f the tertiary sector employment share in the total municipal economically active population. Note: The shades of grey indicate the interpretation of the correlation coefficient based on de Vaus (2002): 0.01-0.09 trivial, no correlation; 0.10-0.29 low; 0.30- 0.49 mild; 0.50-0.69 substantial – strong; 0.70-0.89 very strong; 0.90-1.00 almost perfect. Source: IDOS (2015), Czech Statistical Office (2014a); author’s calculation

149

Tab. 5. Multiple regression analysis of the attractiveness of places (population growth, crude net migration rate and completed dwellings) and socio-economic conditions; municipalities in the South Moravian Region. Variables Unstandardized Standardized Sig. (p Adjusted F Coefficients Coefficients values, two- R2 (Sign. F Dep. Indep. = T B Std. Beta tailed) = 0,000) (car time Error dist.) Constant 1.626 0.026 0.000** 0.245 110.000

SC 1 0.159 0.018 0.378 0.000** SC 2 0.053 0.014 0.164 0.000**

growth Constant 1.618 0.027 0.000** 0.230 201.461 Population Population SC 1 0.202 0.014 0.481 0.000** Constant 1.690 0.028 0.000** 0.213 61.545

SC 1 0.136 0.019 0.320 0.000** SC 2 0.089 0.018 0.270 0.000** SC 3 -0.068 0.017 -0.175 0.000** Crude net Crude net Constant 1.655 0.028 0.000** 0.183 151.648 migration rate migration SC 1 0.018 0.015 0.430 0.000** Constant 1.723 0.026 0.000** 0.166 67.674

SC 1 0.105 0.018 0.268 0.000** SC 2 0.056 0.014 0.185 0.000** **

dwellings Constant 1.715 0.026 0.000 0.146 115.961 Completed Completed SC 1 0.150 0.014 0.384 0.000** Constant 1.584 0.026 0.000** 0.354 123.516 **

SC 2 0.092 0.017 0.270 0.000 SC 1 0.109 0.018 0.247 0.000** SC 3 0.079 0.016 0.197 0.000**

educated ** - University Constant 1.737 0.015 0.000 0.301 289.298 SC 2 0.188 0.011 0.549 0.000** Constant 1.596 0.030 0.000** 0.270 125.364 ** SC 3 0.172 0.016 0.387 0.000 SC 1 0.117 0.017 0.241 0.000**

sector **

in tertiary in tertiary Constant 1.776 0.016 0.000 0.220 190.298 Employment Employment SC 3 0.209 0.015 0.470 0.000** Constant 1.511 0.039 0.000** 0.180 74.869 SC 1 0.233 0.022 0.392 0.000**

SC 3 0.045 0.020 0.082 0.027* rate Constant 1.525 0.039 0.000** 0.176 144.011 ** Unemployment Unemployment SC 1 0.250 0.021 0.421 0.000 * significant at 0.05 level, ** significant at 0.01 level Source: author’s calculation

150

All predictors together explain 35.4 per cent of the variance of the university-educated population. However, access to services within the SC 2 category contributes solely by 30.1 per cent. As the services included in the SC 2 category are typical for medium-sized and larger towns (urban centres with approximately more than 10,000 inhabitants in the South Moravian Region’s case, see Figure 1), access to such services can explain one third of the university- educated population’s close proximity to these centres. Focusing on the population change, access to services creating the largest catchment areas and localized in the biggest centres (SC 1 category) contribute to the variance of population growth by 23.0 per cent. Similarly, 27.0 per cent of the employment in the tertiary sector variability can be explained by the access to services SC 3 and SC 1 (22.0% by SC 3). Other variables have an even worse explanatory power. The unemployment rate, completed dwellings, and crude net migration rate indicators are linked to the access to services SC 1, but explain hardly one fifth of the indicators’ variability. A mild association of variables with each other is reflected in the rather low explanatory power of the T values with regard to the level of socio-economic disadvantage. Although spatial differences in access to SGI exist, their potential impact on the demographic, social and economic situations of the particular municipalities is not strong. Based on the T categorization, the access to services with the highest SC (SC 1) is predominantly the most important predictor. In this context, other factors including economic attractiveness in terms of jobs, local entrepreneurship potential or agglomeration economies, and the social climate within a particular location could play the role of explanatory variables.

5 Concluding remarks According to EU regional policy, access to SGI is related to the spatiality of human lives determined by an uneven distribution of socio-economic activities. The emphasis on the spatial dimension of living standards is best represented by the suggestion of the Third Report on Economic and Social Cohesion that ‘people should not be disadvantaged by wherever they happen to live or work in the Union’ (EC 2004a, p. 27). However, the location of the residence and socially valued resources is one of the fundamental factors contributing to spatial disparities in access to SGI on all geographical scales. The question is whether inadequate access worsens living conditions in the affected places. In this context, a certain limit of differences in access that could be used to evaluate the ‘fairness’ or ‘adequateness’ of SGI accessibility has not yet been clarified. Based on the principles of the Rawlsian ‘theory of justice’ (Lévy, 2003, cited in Peyrony, 2007), a territorially cohesive situation can be achieved when access to SGI does not have a negative impact on the living conditions of particular places. This paper aimed at revealing the relationship between unequal access to SGI and the level of local development (in terms of population ageing, attractiveness of places and socio-economic conditions), using the South Moravian Region in the Czech Republic as a case study. The ‘fairness’ principle contained in the territorial cohesion discourse relativizes the stringent aim of equality and adds the ideas of social solidarity and spatial justice (Soja, 2010) to the economic model of development dominated by market forces. The preservation of public goods by public authorities in disadvantaged regions is a perfect example of a socially motivated attitude towards diminishing regional disparities that Davoudi (2005, p. 435) called

151

‘spatializing the European social model’. However, in the case of the South Moravian Region, the market failure in the field of access to SGI is compensated by the public provision of basic services with varying success. Equal opportunities to access SGI seem to be problematic simply due to the hierarchy of urban systems and its stability over the course of time. In this regard, a proper functioning of public transport and the quality of the transport network should help increase the SGI accessibility. While spatial diversity is not seen as a barrier to development in itself (EC, 2008), declining demographic and socio-economic conditions caused by inadequate access to SGI indicate a situation which is contrary to the territorial cohesion principles. The relative distance to the centres strongly depends on urban hierarchy and transport infrastructure. Even if the municipality is physically close to the centre, inadequate infrastructure (e.g., slip roads) and limited public transport connections (due to a smaller size of municipality) have an impact on the final accessibility. The ‘tunnel effect’ is slightly visible in the case of the motorway (D2) heading south from Brno, although only if public transport connections and fare are taken into account. In this regard, the motorway does not serve as a favourable mode of infrastructure for public transport, which requires flexibility and depends more on timetables and the form of route plans. Conversely, more distant but better connected municipalities (mostly more populated) are characterized by improved access. The effect of the time-space compression is primarily visible when looking at the routing of motorways and railway corridors and the position of the centres. The level of SGI accessibility is determined by the direct access to the transport infrastructure. However, the potential ‘pump effect’ of transport networking increases the loss of the service functions in the non-central municipalities. This is the case for municipalities intersected by major transport infrastructure in the Brno surroundings (e.g., Kuřim and smaller municipalities in the southern part of the Brno metropolitan area). Reflecting the discussion about the agglomeration shadow, municipalities in the Brno hinterland are prone to lose basic services compared to similar sized municipalities located further away from Brno; such a situation deserves more research attention. SGI accessibility is rather uneven in the South Moravian Region. The excluded municipalities are predominantly located in the peripheral areas close to the regional borders (Znojmo area) and in areas lacking a major service centre, i.e. ’inner pockets’ (southeast and north of Brno). These territories suffer from worse transportation accessibility of (occasionally daily-used) services. Compared to the more central areas (in the context of access to SGI), the opportunities to reach services for these less favoured municipalities have deteriorated (and are not equal) due to a higher relative distance to the centres. Such a situation goes hand in hand with the depopulation of several municipalities affected by worse automobile transport SGI accessibility. Finding a way out of this ‘downward spiral’ will not be an easy task. Emigration movements from the periphery localities (Znojmo area, the north-western and eastern part of the region), which have occurred since the beginning of 1990s (Czech Statistical Office, 2014c), are related to the decrease in importance of the medium-sized cities and local rural centres. Since then, the local centres directly supported by the socialist planning policy have partially lost their service function as a result of the competitive pressure originating from the stronger centres (especially Brno), as well as being related to the economic transformation from a socialist to a democratic society. However, the impact of SGI accessibility on the attractiveness

152 of places should not be overestimated. The multiple regression analysis shows only a partial explanation for the relationship between the variables. This is also to a limited extent the case of socio-economic conditions that are affected by unequal access. In light of these empirical results, it cannot be assumed that unequal access to SGI has a strong impact on territorial cohesion in terms of the spatial disparities in living conditions, at least in the municipalities of the South Moravian Region. The variability of the overall level of development can only partially be explained by different levels of accessibility (ranging between one third and one fifth of the dependent variable that can be predicted by different values of accessibility). Besides unequal access to SGI, other factors contribute to the living conditions of particular places. Moreover, the presented analysis does not provide a more detailed picture of the population structure and its relation to SGI accessibility. Consequently, spatial access to services should be complemented by research on travel behaviour regarding socio-economic background, age, gender, culture, traditions, etc. (see e.g., Haugen et al., 2012 focusing on the temporal dimension or Rodriguez and Rogers, 2014), on the qualitative features of services (e.g., capacity of the facility, opening hours, quality of service), and research aimed at the perception of an acceptable level of SGI accessibility (an overview of discourses of ‘access’ is provided by Cass et al., 2005). In this respect, the results of the present study should serve as an initial framework for subsequent research.

153

154

6 CONCLUSIONS

The dissertation focused on territorial cohesion, an umbrella concept covering a wide range of aspects related to territorial development. Although territorial cohesion has a relatively strong position at the political level, mainly because of its official inclusion among the fundamental EU objectives by Lisbon Treaty, the actual meaning remains to a great extent ambiguous (Waterhout, 2008; Servillo, 2010). Such fuzziness, stemming particularly from the political nature of the term seeking only to set the framework for decisions on spatial development (Faludi, 2005a), considerably complicates both the definition/interpretation of territorial cohesion and its analytical grasp from the spatial planning perspective. Non-binding platform of spatial planning at EU level, based on the ESDP as a pivotal document forming the EU position towards spatial development, influences the final form of normative policy by its own recommendations, which are in some views difficult to understand and even hardly compatible with various traditions of spatial planning in member states (Maier, 2004; Faludi, 2010). The result is the coexistence of different approaches operationalizing territorial cohesion as a multidimensional phenomenon (e.g. ESPON 2006a; EU Ministers responsible for Spatial Development, 2011; ESPON, 2012c; Abrahams, 2014). The definition of the partial dimensions, however, is itself based on the thesis that the concept of territorial cohesion is composed of these dimensions and acquires all their meanings. Such a syntactic character of multidimensional perspective is contradicted by the discursive understanding of territorial cohesion, which identifies alternative and competing (sometimes intersecting) interpretations of the concept that form the basic ideological background of territorial cohesion policy (e.g. Waterhout, 2007; Evers et al., 2009). This conflict of thoughts can be found in the dual interpretation of territorial cohesion as a path to balanced territorial development or competitiveness (Waterhout, 2007; Ache et al., 2008; Vanolo, 2010). In this context, the idea of polycentric development became accentuated in recent years. Polycentricity, as a bridging concept, should integrate competing interpretations and alleviate their semantic inconsistency. Polycentricity itself, however, is another elusive concept, whose impact on territorial cohesion has not yet been sufficiently empirically verified. The presented study focused through five objectives (each elaborated in a separate article) on the issue of territorial cohesion – in terms of its relevance and its relation to selected national spatial planning tradition (Paper 1), in terms of the role of polycentric development in the current dynamic development of spatial and functional settings of urban systems (Paper 2 and 3) and in terms of access to basic services as elemental assumption of territorially balanced development and reduction of regional disparities (Paper 4 and 5). Each objective was specified by relevant research questions. Territories of the Czech Republic and the South Moravian Region serve as study areas. The hypothesis about similarities of territorial cohesion discourses of the current EU regional policy and the former socialist planning doctrine was confirmed to some extent. Both spatial planning doctrines highlight the importance of metropolitan areas at regional and national scale and the need to balance their dominant position by supporting rural and peripheral areas - a phenomenon which can be described as a concentration/de-concentration duality. As well as the current vision of balanced Europe, the socialist spatial planning realized the need

155 for an integrated approach to specific spatial units in later years of its operation. The difference can be seen in the specific functional structure within which the relationships between places were comprehended. Especially the organizational structure of urban agglomerations has been viewed from a different perspective. While the central socialist planning perceived the spatial arrangement of individual spatial units as hierarchical with rather one-sided and vertical linkages, the current concept of ITI reflects the natural dynamic of metropolitan regions, which lies more in the horizontal configuration of spatial relationships, deconcentration processes, and specialization of centres. The role smaller centres in peripheral areas play for the development of particular regions is considered crucial by both doctrines. However, understanding of their function, i.e. the mode of production, within the broader urban system differs. The difference in perspective arises from the differences of the ideological frameworks (communism/democracy) and economic systems (collectivization/free market economy). Although the idea of balanced spatial development is, by its nature, close to the equalizing policy of the socialist regime, its application by the centrally managed apparatus of spatial planning was in practice often ineffective. On the contrary, the current territorial cohesion discourse stresses protection of the specifics of individual regions as one of the key aspects of regional development potential. In the capitalist environment of the current EU, however, ambitions of territorial cohesion seem to be rather implausible and idealistic. Deeper insight into the development of relational space hierarchy revealed that normative representations of spaces indeed largely differ from the real spatial organization and dynamics of routine activities, at least in the Czech Republic. The former spatial configuration of the Czech urban system tending to take the form of polycentric arrangement due to the applied concept of ‘central settlement system’ (but with very questionable results) has been affected by the competitive environment and the spontaneity of socioeconomic transformation in the last 25 years. The space(s) of everydayness, analysed through the dynamic of labour systems, has experienced relatively intensive changes in the last decades. The role of central places in organization and representation of space is increasing, at least in regard to the regional (and also national) level. Consequently, re-centralization of relational space characterized by a change of scale and spatial embeddedness of daily practices can be observed. The analysis showed a loss of significance of local job centres and pointed out the growing importance of the largest agglomerations. Increasing spatial polarization is therefore in contrast with the ‘grand narratives’ of the current cohesion policy. However, even in the Czech territory, regions with prevailing polycentric arrangement can be identified (rather peripheral regions with medium-sized and small towns). This information is important for answering the question whether the polycentric structure of an urban system actually contributes to reducing spatial disparities, as territorial cohesion discourse proclaims. The analysis aimed at the relation between spatial disparities and the degree of polycentricity did not find the close association (with the exceptions of certain indicators and size categories). This micro-regional analysis together with the results of researches focused on the regional and international scale (Meijers and Sandberg, 2006; Veneri and Burgalassi, 2012) can serve as a basis for rejecting the hypothesis that a blind faith in the efficiency of polycentric development for acceleration of spatial development is empirically proofed.

156

The territorial cohesion discourse is closely tied with the issue of access to basic services, labelled as services of general interest (SGI) by the cohesion policy terminology. Evenness of their provision and, if possible, their fair accessibility, is one of the basic priorities of EU balanced spatial development. In this context, however, the accessibility of SGI is strongly influenced by functioning of the urban system. The reason is the simple fact that SGI are distributed unevenly and especially residents of the smallest municipalities in peripheral locations are dependent on the often lengthy commuting. In the Czech urban system, the service function of small towns seems to be crucial. The logic of central places, however, applies to the service function of small towns only partially. Although, with some degree of generalization, the supply of services increases with the population size of settlements, there is considerable heterogeneity in the degree of service function with regard to particular size categories (i.e. cities of similar size). The key reason is the varying degree of peripherality of small towns given by population density and proximity to the nearest larger centre. Small towns in peripheral locations have greater service function than similarly sized towns situated in the hinterland of larger centres and especially the largest agglomerations. In this regard, small towns are affected by the ‘agglomeration shadow’, i.e. the proximity of a centre which competes with the small town and has a negative effect on its service function. While spatial position is a very important aspect for the actual provision of services (morphological perspective), the degree of functional integration of a small town into the broader urban system through transport infrastructure is almost irrelevant – functionally integrated small towns do not offer more services than those not integrated, as the concept of ‘borrowed size’ assumes (with the exception of the axis of commercial suburbanization south of Brno). In this context, there are some easily accessible local centres characterized by only average supply of services and more local character of consumer behaviour. Higher occurrence of greater service function is typical for small towns that are attractive for tourists. In this case, the degree of service function does not relate to the geographic position of the town in the urban system, but partly with something that is called ‘inner potential’. The final part of the study worked with the thesis that unequal access to SGI is one of the causes of social and spatial disparities. Unequal access, stemming mainly from the irregular spatial presence of urban centres and inaccessible transport infrastructure in some areas, however, is an integral part of spatial reality and not to be, a priori, considered a factor contributing to territorial inconsistency. SGI accessibility is thus related to the issues of social solidarity and spatial justice. Access to SGI has been labelled insufficient or unfair when contributing to the deterioration of socioeconomic conditions of a given area or place. Although spatial inequalities in living conditions certainly exist, the study showed that their occurrence is associated with varying levels of access to SGI only minimally. The concentration of the university-educated and the employed in the service sector may be partly explained by the higher level of SGI accessibility. The relationship is even weaker for other variables, suggesting that socioeconomic differences are probably associated with other aspects that were not taken into account in the analysis. The territorial cohesion discourse is supported by political establishment for its principally noble intentions. At the spatial planning level, it is very difficult to operationalize such a complex notion. Using the concept of polycentricity and SGI accessibility, the meaning

157 of the universally accepted ‘buzzword’, which the term ‘territorial cohesion’ undoubtedly became, should be further clarified. However, the polycentric development normative, as a tool for achieving territorial cohesion, diverges from developmental dynamics of the urban system and is not directly associated with levels of spatial inequalities. The growing polarization of economic activities and the multifaceted functional processes taking place within the urban networks greatly complicate not only the accessibility of jobs but also SGI. In this respect, it is evident that the growing role of the largest agglomerations has a devastating effect on urban functions of municipalities in their hinterland. In spite of that, SGI accessibility persists as a problem, especially in peripheral areas. As demonstrated by the presented dissertation, excessive generalizations, and simplistic interpretations must be avoided, because of the particularities and uniqueness of individual regions and places. The spatial reality is based on inequality, and homogeneous space is only an idea and a vision. Similarly, territorial cohesion is only a representation of space, which does not reflect the diversity of natural spatial arenas where everyday human activities occur.

158

7 ZÁVĚRY

Dizertační práce se zabývala územní soudržností (‘territorial cohesion’), zastřešujícím konceptem pokrývajícím široké spektrum aspektů týkajících se územního rozvoje. Ačkoliv má koncept územní soudržnosti v politické rovině poměrně silné postavení, a to především z důvodu oficiálního začlenění územní soudržnosti mezi základní cíle EU Lisabonskou smlouvou, jeho skutečný význam zůstává do vysoké míry nejednoznačný (Waterhout, 2008; Servillo, 2010). Tato nejasnost, pramenící zejména z politické povahy konceptu usilující pouze o nastavení rámce v oblasti rozhodování o územním rozvoji (Faludi, 2005a), výrazně ztěžuje jak samotnou definici a interpretaci územní soudržnosti, tak její analytické uchopení v rovině prostorově plánovací. Nezávazná platforma prostorového plánování na úrovni EU, vycházející z ESDP jako stěžejního dokumentu formulujícího postoj EU vůči územnímu rozvoji, pak svými doporučeními ovlivňuje výslednou podobu politicky závazných normativů, které jsou však v některých pohledech těžko pochopitelné, a dokonce obtížně slučitelné s tradicemi prostorového plánování jednotlivých členských států (Maier, 2004; Faludi, 2010). Výsledkem je koexistence různých přístupů operacionalizujících územní soudržnost jako multidimenzionální jev (např. ESPON 2006a; EU Ministers responsible for Spatial Development, 2011; ESPON, 2012c; Abrahams, 2014). Samotné vymezování dílčích dimenzí však již samo o sobě vychází z teze, že koncept územní soudržnosti je z těchto dimenzí složen a nabývá všech jejich významů. Tento skladebný charakter multidimenzionální perspektivy je však rozporován diskurzivním chápáním územní soudržnosti, které identifikuje alternativní a vzájemně soupeřící (i prolínající se) interpretace konceptu tvořící základní ideové pozadí politiky územní soudržnosti (např. Waterhout, 2007; Evers et al., 2009). Klíčový myšlenkový střet lze nalézt v duální interpretaci územní soudržnosti jako cestě k vyváženému územnímu rozvoji nebo konkurenceschopnosti (Waterhout, 2007; Ache et al., 2008; Vanolo, 2010). V této souvislosti se v posledních letech dostala do popředí zájmu myšlenka polycentrického rozvoje (‘polycentric development’), který by měl jako propojující koncept (‘bridging concept‘) vzájemně soupeřící interpretace integrovat a mírnit jejich významový nesoulad. Samotná polycentricita je však dalším poměrně obtížně uchopitelným konceptem, jehož význam pro územní soudržnost nebyl dostatečně empiricky verifikován. Prezentovaná studie se prostřednictvím pěti dílčích cílů (každý rozpracován v samostatném článku) zaměřila na problematiku územní soudržnosti z hlediska její relevance a vztahu k vybrané národní tradici územního plánování (Článek 1), z hlediska významu polycentrického rozvoje v současném dynamickém vývoji prostorových a funkčních vztahů formujících sídelní systémy (Článek 2 a 3) a z hlediska dostupnosti základních služeb jakožto elementárního předpokladu územně vyváženého rozvoje a mírnění vzniku územních disparit (Článek 4 a 5). Každý dílčí cíl byl upřesněn výzkumnými otázkami. Případovou studií se stalo území České republiky s důrazem na Jihomoravský kraj. Hypotéza o podobných rysech diskurzů územní soudržnosti současné regionální politiky EU a dřívější doktríny socialistického plánování se do jisté míry potvrdila. Oba prostorově plánovací přístupy vyzdvihují jak důležitost metropolitních oblastí v regionálním i národním měřítku, tak potřebu vyvažování jejich dominantní funkce podporou rurálních a periferních oblastí - jev, který lze označit jako koncentrační/dekoncentrační dualitu

159

(‘concentration/de-concentration duality’). Stejně jako současný diskurz vyvážené Evropy, tak i socialistické prostorové plánování si v pozdějších dobách svého fungování uvědomilo nutnost integrovaného přístupu ke specifickým prostorovým jednotkám. Rozdíl lze spatřit v konkrétní funkční struktuře, v jaké byly vztahy mezi místy nahlíženy. Především organizační uspořádání městských aglomerací bylo nazíráno odlišnou optikou. Zatímco centrální socialistické plánování vnímalo prostorové uspořádání jednotlivých jednotek jako hierarchické se spíše vertikálními a jednostrannými vazbami, současný koncept ITI reflektuje přirozenou dynamiku metropolitních regionů, která spočívá více v horizontální konfiguraci prostorových vazeb, dekoncentračních procesech a specializaci sídel. Role menších středisek v periferních oblastech pro rozvoj příslušných regionů je oběma doktrínami vnímána jako zásadní. Liší se však pohled na jejich funkci, resp. mód produkce, v rámci širšího sídelního systému. Rozdílná perspektiva vyplývá z odlišného ideologického rámce (komunismus/demokracie) a ekonomického systému (kolektivizace/volný trh). Ačkoliv je myšlenka vyváženého územního rozvoje svou podstatou bližší nivelizační politice socialistického režimu, její aplikace centrálně řízeným aparátem prostorového plánování v praxi byla v mnoha případech neefektivní. Naopak současný diskurz územní soudržnosti vyzdvihuje ochranu specifik jednotlivých regionů, na kterých by měl být rozvoj vystavěn. V kapitalistickém prostředí současné EU však působí ambice konceptu územní soudržnosti nevěrohodně až idealisticky. Při hlubším pohledu na vývoj hierarchie relačního prostoru bylo zjištěno, že normativní reprezentace prostorů se skutečně z velké části odlišují od reálné prostorové organizace společnosti a dynamiky rutinních pohybů, alespoň v rámci České republiky. Bývalá prostorová konfigurace sídelního systému, tíhnoucí k polycentrickému uspořádání v důsledku aplikovaného konceptu střediskové soustavy osídlení (avšak s velmi diskutabilními výsledky), je v posledních zhruba 25 letech ovlivňována konkurenčním prostředím a spontánností socioekonomické transformace. Prostor(y) každodennosti (‘space(s) of everydayness’), analyzován prostřednictvím vývojové dynamiky pracovních trhů, tudíž zažívá v posledních dekádách poměrně intenzivní změny. Stále větší roli v organizaci a reprezentaci prostoru hrají centrální místa, alespoň co se týče regionální (potažmo národní) úrovně. Postupně tak dochází k re-centralizaci relačního prostoru vyznačující se změnou měřítka a prostorovým zakořeněním každodenních praktik. Analýza prokázala značnou ztrátu významnosti lokálních pracovních center, a naopak poukázala na rostoucí důležitost největších aglomerací. Rostoucí polarizace území je tak v silném rozporu s velkými narativy (‘grand narratives’) současné kohezní politiky. I přesto lze v českém prostoru identifikovat regiony s převládajícím polycentrickým uspořádáním (spíše periferní regiony se středně velkými a malými městy). Tato informace je důležitá při zodpovězení otázky, zda polycentrická struktura sídelního systému skutečně přispívá ke zmenšení územních disparit tak, jak je diskurzem územní soudržnosti proklamováno. Vlastní analýza vztahu míry územních disparit a míry polycentricity však těsnou souvislost nenašla (až na výjimky některých ukazatelů a velikostních kategorií regionů). Tato mikro-regionální analýza spolu s výsledky výzkumů zaměřených na regionální až mezinárodní měřítko (Meijers and Sandberg, 2006; Veneri and Burgalassi, 2012) slouží jako podklad pro odmítnutí teze, že víra v samo-spásnou úlohu polycentricity pro akceleraci vyváženého územního rozvoje je empiricky podložena.

160

Diskurz územní soudržnosti je od počátku úzce svázán s problematikou dostupnosti základních služeb, terminologií kohezní politiky označovaných jako služby obecného zájmu (‘services of general interest - SGI’). Rovnoměrnost jejich nabídky a pokud možno jejich spravedlivá dostupnost patří mezi základní priority územně vyváženého rozvoje EU. V této souvislosti je však právě dostupnost SGI silně ovlivněna fungováním sídelního systému. Důvodem je prostý fakt, že SGI jsou rozmístěny nerovnoměrně, a tak především obyvatelé nejmenších sídel v periferních lokalitách jsou odkázáni na mnohdy zdlouhavou dojížďku. V rámci českého sídelního systému je v tomto ohledu klíčová obslužná funkce malých měst. Logika centrálních míst však platí pro obslužnou funkci malých měst pouze částečně. Ačkoliv lze s jistou mírou generalizace tvrdit, že nabídka služeb se zvyšuje spolu s populační velikostí sídla, v rámci velikostních kategorií (tedy velmi podobně velkých měst) existuje značná heterogenita v míře obslužné funkce. Kruciálním důvodem je míra perifernosti malých měst daná hustotou zalidnění a blízkostí nejbližšího většího centra. Malá města v periferních polohách disponují vyšší obslužnou funkcí než přibližně stejně velká malá města situovaná v zázemí větších center, a především největších aglomerací. V této souvislosti jsou malá města ovlivněna „aglomeračním stínem“ (‘agglomeration shadow’), tedy blízkostí konkurenčního střediska - to přináší silnou konkurenci malému městu a v důsledku má negativní vliv na jeho obslužnou funkci. Zatímco z hlediska morfologického je poloha velmi důležitým aspektem pro výslednou nabídku služeb, míra funkčního zapojení malého města do širšího sídelního systému prostřednictvím dopravní infrastruktury téměř nehraje roli – funkčně integrovaná malá města nenabízí více služeb než ta neintegrovaná, jak předpokládá koncept „půjčené velikosti“ (‘borrowed size’); výjimku tvoří osa komerční suburbanizace jižně od Brna. V tomto kontextu jsou i některá dobře dostupná lokální střediska charakteristická pouze průměrnou nabídkou služeb a spíše lokálním charakterem spotřebního chování. Vyšší pravděpodobnost větší nabídky služeb je pak typická pro turisticky atraktivní malá města. V tomto případě nesouvisí míra obslužná funkce s polohou sídla v sídelním systému, ale částečně s něčím, co je označeno jako „vnitřní potenciál“ (‘inner potential’). Závěrečná část studie pracovala s tezí, že nerovný přístup k SGI je jedním z původců sociálně prostorových disparit. Nerovnoměrný přístup, pramenící zejména z nepravidelného výskytu městských center a v některých prostorech nedostatečného napojení na dopravní infrastrukturu, je ovšem nedílnou součástí prostorové reality a nelze být a priori považován za faktor přispívající k územní nesoudržnosti. Dostupnost SGI je tak spojena s otázkami sociální solidarity a prostorové spravedlnosti. Dostupnost SGI byla označena jako nedostatečná nebo nespravedlivá v případě, kdy přispívá ke zhoršení socioekonomických podmínek dané oblasti či místa. Ačkoliv prostorové nerovnosti v životních podmínkách rozhodně existují, studie prokázala, že je jejich výskyt v pouze malé míře asociován s rozdílnou mírou dostupnosti SGI. Koncentraci vysokoškolsky vzdělaných a zaměstnaných ve službách lze částečně vysvětlit vyšší mírou dostupnosti SGI. U ostatních proměnných je vztah ještě menší, z čehož vyplývá, že socioekonomické rozdíly budou mít souvislost s dalšími aspekty, které nebyly analýzou zohledněny. Diskurz územní soudržnosti je politickou reprezentací podporován pro své principiálně šlechetné úmysly. V rovině prostorově plánovací je však velmi složité takto komplexní termín operacionalizovat. Pomocí konceptů polycentricity a dostupnosti SGI by měl být objasněn

161 obsah všeobecně akceptovaného módního slova (‘buzzword’), kterým se termín územní soudržnost bezesporu stal. Samotný normativ polycentrického rozvoje, jako nástroje dosahování územní soudržnosti, se však zásadně rozchází s vývojovou dynamikou osídlení a nemá přímou souvislost s mírami územních nerovností. Rostoucí polarizace ekonomických aktivit a mnohovrstevnatost funkčních procesů v rámci městských sítí výrazně komplikuje nejen problematiku dostupnosti pracovních míst, ale také SGI. I v tomto ohledu je zřejmá sílící role největších aglomerací, které „vysávají“ městské funkce sídlům ve svém zázemí. I přesto je dostupnost SGI přetrvávajícím problémem hlavně v periferních oblastech. Jak ale ukázala předkládaná disertační práce, přílišným generalizacím a zjednodušujícím interpretacím je potřeba se vyhýbat, a to z důvodu specifik a jedinečností jednotlivých regionů a míst. Prostorová realita je založena na nerovnostech a homogenní prostor je pouze ideou a představou. Podobně územní soudržnost představuje pouze reprezentaci prostoru, která plně nereflektuje přirozenou různorodost prostorových arén každodenních lidských aktivit.

162

REFERENCES

ABRAHAMS, G. (2014): What “Is” Territorial Cohesion? What Does It “Do”?: Essentialist Versus Pragmatic Approaches to Using Concepts. European Planning Studies, 22 (10), pp. 2134–2155. ACHE, P., ANDERSEN, H. T., MALOUTAS, T., RACO, M., TASAN-KOK, T. ed. (2008): Cities between Competitiveness and Cohesion: Discourses, Realities and Implementation. Springer, Berlin, 306 p. AHNER, D. (2009): What do you really know about European cohesion policy?, http://economia.unipr.it/DOCENTI/WOLLEB/docs/files/Ahner.pdf (16/1/2014). ALBRECHTS, L. (2001): How to proceed from image and discourse to action: As applied to the Flemish Diamond. Urban Studies, 38 (4), pp. 733–745. ALLEN, D. (2005): Cohesion and the Structural Funds: Competing Pressures for Reform? In: WALLACE, H., WALLACE, W., POLLACK, M. A. ed.: Policy-Making in the European Union. Oxford University Press, Oxford, pp. 213–241. ALMUNIA, J. (2011): Reform of the State aid rules for services of general economic interest (SGEI) and decisions on WestLB, Bank of Ireland and France Telecom, http://www.europa.eu/rapid/press-release_SPEECH-11-901_en.pdf (29/4/2015). ALONSO, W. (1973): Urban zero population growth. Daedalus, 102 (4), pp. 191–206. ANAS, A., ARNOTT, R., SMALL, K. A. (1998): Urban spatial structure. Journal of Economic Literature, 36 (3), pp. 1426–1464. BAILLY, A. (2001): Conclusions: Europe, today, tomorrow. In: BAILLY, A., FREMONT, A. ed.: Europe and Its States: A Geography. DATAR, Paris, pp. 195–197. BANKY (2015): banky.cz, http://www.banky.cz/ (3/3/2015). BATTEN, D. F. (1995): Network cities: creative urban agglomerations for the 21st century. Urban Studies, 32 (2), pp. 313–327. BERRY, B. J. L. (1964): Cities as systems within systems of cities. Papers of the Regional Science Association, 13 (1), pp. 146–163. BERRY, B. J. L., PARR, J. B. (1988): Market Centres as Retail Locations. Prentice Hall, Englewood Cliffs, NJ, 230 p. BERRY, B. J. L., PRED, A. (1965): Central Place Studies: A Bibliography of Theory and Applications. Regional Science Research Institute, Philadelphia, 203 p. BLAIR, A. (2005): The European Union since 1945. Pearson, Harlow, 159 p. BLAŽEK, J. (2008): Teorie regionálního rozvoje. In: WOKOUN, R., MALINOVSKÝ, J., DAMBORSKÝ, M., BLAŽEK, J. et al. ed.: Regionální rozvoj: východiska regionálního rozvoje, regionální politika, teorie, strategie a programování. Linde, Praha, p. 220–281. BLAŽEK, J., UHLÍŘ, D. (2011): Teorie regionálního rozvoje: nástin, kritika, implikace. Karolinum. Karolinum, Praha, 342 p.

163

BOHÁČKOVÁ, I., HRABÁNKOVÁ, M. (2009): Strukturální politika Evropské unie. C. H. Beck, Praha, 188 p. BÖHME, K., DOUCET, P., KOMORNICKI, T., ZAUCHA, J., ŚWIĄTEK, D. (2011): How to strengthen the territorial dimension of Europe 2020 and the EU Cohesion Policy. Warsaw. BRET, B. (2009): Rawlsian universalism confronted with the diversity of reality. justice spatiale|spatial justice, 1, 7 p. BROTCHIE, J., BATTY, M., BLAKELEY, E., HALL, P., NEWTON, P. ed. (1995): Cities in Competition. Longman, Melbourne, 496 p. BRUNET, R. (1989): Les Villes Europe´ennes: rapport pour la DATAR. La Documentation Francaise, Paris. BURGER, M., MEIJERS, E. (2012): Form follows function? Linking morphological and functional polycentricity. Urban Studies, 49 (5), pp. 1127–1149. BURGER, M. J., MEIJERS, E. J., HOOGERBRUGGE, M. M., TRESSERRA, J. M. (2015): Borrowed Size, Agglomeration Shadows and Cultural Amenities in North-West Europe. European Planning Studies, 23 (6), pp. 1090–1109. BURGER, M. J., VAN DER KNAAP, B., WALL, R. S. (2014): Polycentricity and the multiplexity of urban networks. European Planning Studies, 22 (4), pp. 816–840. BURGESS, E. W. (1925): The growth of the city: An introduction to a research project. In: PARK, R. E., BURGESS, E. W., McKENZIE R. D. ed.: The City. University of Chicago Press, Chicago, pp. 47–62. BURNHAM, J. (1999): Contrasts and contradictions in French regional policy: from de Gaulle to Chirac, from the Treaty of Rome to the Treaty of Amsterdam, from central direction to local cooperation. In: ALLISON, M., HEATHCOTE, O. ed.: Forty Years of the Fifth French Republic: Actions, Dialogues and Discourses. Peter Lang, Bern, pp. 77–94. BURTON, I. (1963): A restatement of the dispersed city hypothesis. Annals of the Association of American Geographers, 53 (3), pp. 285–289. CAMAGNI, R. (1993): From city hierarchy to city network: Reflections about an emerging paradigm. In: LAKSHMANAN, T. R., NIJKAMP, P. ed.: Structure and Change in the Space Economy. Springer Verlag, Berlin, pp. 66–87. CAMAGNI, R. (2006): Territorial Impact Assessment - TIA: A methodological proposal. Scienze Regionali - Italian Journal of Regional Science, 5 (2), pp. 135–146. CAMAGNI, R. (2007): Territorial Development Policies in the European Model of Society. In: FALUDI, A. ed.: Territorial Cohesion and the European Model of Society. Lincoln Institute of Land Policy, Cambridge, Massachusetts, pp. 129–144. CAMAGNI, R. (2009): Territorial Impact Assessment for European regions: A methodological proposal and an application to EU transport policy. Evaluation and Program Planning, 32 (4), pp. 342–350.

164

CAMAGNI, R. P., SALONE, C. (1993): Network Urban Structures in Northern Italy: Elements for a Theoretical Framework. Urban Studies, 30 (6), pp. 1053–1064. CAPELLO, R. (2000): The city network paradigm: Measuring urban network externalities. Urban Studies, 37 (11), pp. 1925–1945. CASS, N., SHOVE, E., URRY, J. (2005): Social exclusion, mobility and access. The Sociological Review, 53 (3), pp. 539–555. CASTELLS, M. (1996): The Rise of the Network Society. 2nd ed. Blackwell, Oxford, 594 p. CEBOLLADA, A. (2009): Mobility and labour market exclusion in the Barcelona Metropolitan Region. Journal of Transport Geography, 17 (3), pp. 226–233. CEC (1999): ESDP - European Spatial Development Perspective: Towards Balanced and Sustainable Development of the Territory of the European Union. Office for Official Publications of the European Communities, Luxembourg, 82 p. CEEP (2010): Mapping of the Public Services, May 2010, http://www.ceep.eu/images/stories/pdf/Mapping/CEEP_mapping%20experts%20report.pdf (12/8/2015). CHRISTALLER, W. (1933): Die Zentralen Orte in Süddeutschland. Gustav Fischer, Jena, 331 p. CHURCH, A., REID, P. (1999): Cross-border Co-operation, Institutionalization and Political Space Across the English Channel. Regional Studies, 33 (7), pp. 643–655. CIGALE, D., LAMPIČ, B., OGRIN, M., PLUT, D., REBERNIK, D., ŠPES, M., VINTAR MALLY, K., CETKOVSKÝ, S., KALLABOVÁ, E., MIKULÍK, O., VAISHAR, A., ZAPLETALOVÁ, J. (2006): Sustainable development of small towns a Slovenian-Moravian comparative methodological approach. Moravian Geographical Reports, 14 (1), pp. 17–28. CLARK, W. A. V., KUIJPERS-LINDE, M. (1994): Commuting in restructuring urban regions. Urban Studies, 31 (3), pp. 465–483. CLAVAL, P. (1980): Centre-periphery and space: models of political geography. In: GOTTMANN, J. ed.: Centre and periphery: spatial variation in politics. SAGE, Beverly Hills, pp. 63–71. COFFEY, W. J., BOURNE, L. S., RANDALL, J. E., DAVIES, W. K. D., WHITE, R. (1998): Urban systems research: Past, present and future: A panel discussion. Canadian Journal of Regional Science, 21 (3), pp. 327–364. COOMBES, M. (2002): Travel to work areas and the 2001 census. Report to the Office of National Statistics, Newcastle University, Centre for Urban and Regional Development Studies. COPUS, A. K. (2001): From Core-periphery to polycentric development: Concepts of spatial and aspatial peripherality. European Planning Studies, 9 (4), pp. 549–552. COR (2005): Opinion of the Committee of the Regions of 23 February 2005 on the white paper on services of general interest, ECOS/040 – CdR 327/2004 final, 15 March.

165

COTELLA, G., ADAMS, N., NUNES, R. J. (2012): Engaging in European Spatial Planning: A Central and Eastern European Perspective on the Territorial Cohesion Debate. European Planning Studies, 20 (7), pp. 1197–1220. COURTNEY, P., MAYFIELD, L., TRANTER, R., JONES, P., ERRINGTON, A. (2007): Small towns as 'sub-poles' in English rural development: Investigating rural-urban linkages using sub-regional social accounting matrices. Geoforum, 38 (6), pp. 1219–1232. COX, K., MAIR, A. (1988): Locality and Community in the Politics of Local Economic Development. Annals of the Association of American Geographers, 78 (2), pp. 307–325. CRR MU (2009): Polyreg: roční zpráva 2008. CRR MU, Brno. CZECH POST (2015): Search criteria for finding a post office, https://www.postaonline.cz/en/vyhledani-psc (10/3/2015). CZECH STATISTICAL OFFICE (1991): Census 1991, data on commuting. CZECH STATISTICAL OFFICE (2001): Census 2001, data on commuting. CZECH STATISTICAL OFFICE (2011): Census 2011, data on commuting. CZECH STATISTICAL OFFICE (2013): MOS - Municipal statistics, https://vdb.czso.cz/mos/ (1/3/2015). CZECH STATISTICAL OFFICE (2014a): Census 2011, http://www.scitani.cz (21/1/2015). CZECH STATISTICAL OFFICE (2014b): Databáze demografických údajů za obce ČR, http://www.czso.cz/cz/obce_d/index.htm (21/1/2015). CZECH STATISTICAL OFFICE (2014c): Regional Office Brno, https://www.czso.cz/csu/xb (17/7/2015). CZECH STATISTICAL OFFICE (2016): Public database, https://vdb.czso.cz/vdbvo2/ (2/2/2016). CZECHPOINT (2015): CzechPoint, http://www.czechpoint.cz/wwwstats/f?p=100:16 (10/3/2015). ČERMÁK, Z., HAMPL, M., MÜLLER, J. (2009): Současné tendence vývoje obyvatelstva metropolitních areálů v Česku: Dochází k významnému obratu? Geografie – Sborník ČGS, 114 (1), pp. 37–51. DABROWSKI, M. (2014): EU cohesion policy, horizontal partnership and the patterns of sub- national governance: Insights from Central and Eastern Europe. European Urban and Regional Studies, 21 (4), pp. 364–383. DALVI, M. Q., MARTIN, K. M. (1976): The measurement of accessibility: Some preliminary results. Transportation, 5 (1), pp. 17–42. DAVOUDI, S. (2003): Polycentricity in European spatial planning: From an analytical tool to a normative agenda. European Planning Studies, 11 (8), pp. 979–999. DAVOUDI, S. (2005): Understanding Territorial Cohesion. Planning, Practice & Research, 20 (4), pp. 433–441.

166

DAVOUDI, S. (2007): Territorial Cohesion, the European Social Model and Spatial Policy Research. In: FALUDI, A. ed.: Territorial Cohesion and the European Model of Society. Lincoln Institute of Land Policy, Cambridge, Massachusetts, pp. 81–101. DE VAUS, D. (2002): Surveys in Social Research. Allen & Unwin, Crows Nest, 379 p. DINGSDALE, A. (1999): New Geographies of Post-Socialist Europe. The Geographical Journal, 165 (2), pp. 145–153. DOMALEWSKI, P., BAXA, J. (2015): The development of regional differentiation of office construction in the Czech Republic: 1990–2010. Moravian Geographical Reports, 23 (1), pp. 21–33. DOUCET, P. (2006): Territorial Cohesion of Tomorrow: A Path to Cooperation or Competition? European Planning Studies, 14 (10), pp. 1473–1485. EC (1957): Treaty establishing the European Economic Community, http://www.ab.gov.tr/files/ardb/evt/1_avrupa_birligi/1_3_antlasmalar/1_3_1_kurucu_antlasm alar/1957_treaty_establishing_eec.pdf (14/12/2015). EC (1986): The Single European Act. Luxembourg, 23 p. EC (1991): Europe 2000: Outlook for the Development of the Community’s Territory. Brussels, 208 p. EC (1992): Treaty on European Union. Office for Official Publications of the European Communities, Luxembourg, 253 p. EC (1994): Europe 2000+: Cooperation for European territorial development. Office for Official Publications of the European Communities, Brussels, 245 p. EC (1997): Treaty of Amsterdam. Office for Official Publications of the European Communities, Luxembourg, 144 p. EC (1999): ESDP - European Spatial Development Perspective: Towards Balanced and Sustainable Development of the Territory of the European Union. Office for Official Publications of the European Communities, Luxembourg, 82 p. EC (2001): Second report on economic and social cohesion. Publications Office of the European Union, Luxembourg, 158 p. EC (2003): Green paper on services of general interest, http://aei.pitt.edu/45893/1/com2003_0270.pdf (11/4/2015). EC (2004a): Third report on economic and social cohesion. Publications Office of the European Union, Luxembourg, 248 p. EC (2004b): White Paper on services of general interest, http://eur- lex.europa.eu/LexUriServ/site/en/com/2004/com2004_0374en01.pdf (20/10/2015). EC (2006): Community strategic guidelines on economic, social and territorial cohesion, 2007- 2013. Official Journal of the European Union, L 291, pp. 14–32.

167

EC (2007a): Fourth report on economic and social cohesion. Publications Office of the European Union, Luxembourg, 222 p. EC (2007b): Treaty of Lisbon amending the Treaty on European Union and the Treaty establishing the European Community, signed at Lisbon, 13 December 2007. Official Journal of the European Union, C 306, pp. 1–271. EC (2008): Green Paper on Territorial Cohesion: Turning territorial diversity into strength. Brussels, 12 p. EC (2010a): Europe 2020: A strategy for smart, sustainable and inclusive growth. Brussels, 32 p. EC (2010b): Fifth Report on Economic, Social and Territorial Cohesion. Publications Office of the European Union, Luxembourg, 264 p. EC (2012): Consolidated versions of the Treaty on European Union and the Treaty on the Functioning of the European Union. Official Journal of the European Union, C 326, pp. 1–390. EC (2014): Sixth Report on Economic, Social and Territorial Cohesion. Publications Office of the European Union, Luxembourg, 296 p. EISING, R., KOHLER-KOCH, B. (1999): Governance in the European Union: a comparative assessment. In: KOHLER-KOCH, B., EISING, R., ed.: The Transformation of Governance in the European Union. Routledge, London, pp. 267–285. ENYEDI, G. (1990): New Basis for Regional and Urban Policies in East-Central Europe. Discussion Papers, Centre for Regional Studies of Hungarian Academy of Sciences, No. 9. ENYEDI, G. (1996): Urbanization under socialism. In: ANDRUSZ, G., HARLOE, M., SZELENYI, I. ed.: Cities after socialism. Urban and regional change and conflict in post- socialist societies. Willey-Blackwell, Oxford, pp. 100–118. ESPON (2005a): ESPON Project 1.1.1: Potentials for polycentric development in Europe: Project Report. ESPON Programme, Luxembourg, 348 p. ESPON (2005b): ESPON Project 3.2: Spatial Scenarios and Orientations in relation to the ESDP and Cohesion Policy: Second Interim Report. ESPON Programme, Luxembourg, 601 p. ESPON (2006a): ESPON Project 3.2: Spatial Scenarios and Orientations in relation to the ESDP and Cohesion Policy: Final Report. ESPON Programme, Luxembourg, 227 p. ESPON (2006b): ESPON Project 4.1.3: Feasibility Study on Monitoring Territorial Development Based on ESPON Key Indicators, Interim Report. ESPON Programme, Luxembourg, 288 p. ESPON (2007): Scenarios on the territorial future of Europe: ESPON Project 3.2 Final Report. ESPON Programme, Luxembourg, 66 p. ESPON (2008): TIP TAP: Territorial Impact Package for Transport and Agricultural Policies, Inception Report. ESPON Programme, Luxembourg, 77 p. ESPON (2012a): INTERCO: Indicators of territorial cohesion: Final Report - Part B. ESPON & University of Geneva, Luxembourg, 62 p.

168

ESPON (2012b): INTERCO: Indicators of territorial cohesion: Final Report - Part C. ESPON & University of Geneva, Luxembourg, 526 p. ESPON (2012c): KITCASP: Key Indicators for Territorial Cohesion and Spatial Planning, Interim Report. ESPON Programme, Luxembourg, 55 p. ESPON (2013): SeGI: Indicators and perspectives for services of general interest in territorial cohesion and development. ESPON & Royal Institute of Technology, Luxembourg, 57 p. ESPON (2014a): ESPON About, http://www.espon.eu/main/Menu_About/ (24/1/2014). ESPON (2014b): Making Europe Open and Polycentric. ESPON Programme, Luxembourg, 23 p. EU Ministers responsible for Spatial Development (2007): Territorial Agenda of the European Union. Agreed on 24/25 May 2007, Leipzig, http://www.bmub.bund.de/fileadmin/Daten_BMU/Download_PDF/Nationale_Stadtentwicklu ng/territoriale_agenda_en_bf.pdf (14/12/2015). EU Ministers responsible for Spatial Development (2011): Territorial Agenda of the European Union 2020. Agreed on 19 May 2011, Gödöllő, http://www.eu2011.hu/files/bveu/documents/TA2020.pdf (29/12/2015). EVERS, D. (2008): Reflections on Territorial Cohesion and European Spatial Planning. Tijdschrift voor Economische en Sociale Geografie, 99 (3), pp. 303–315. EVERS, D., TENNEKES, J., BORSBOOM, J., HELINGENBERG, H., THISSEN, M. (2009): A Territorial Impact Assessment of Territorial Cohesion for the Netherlands. Netherlands Environmental Assessment Agency (PBL), Hague, 71 p. EZCURRA, R., GIL, C., PASCUAL, P., RAPÚN, M. (2005): Inequality, polarisation and regional mobility in the European Union. Urban Studies, 42 (7), pp. 1057–1076. FALUDI, A. (2004): Territorial Cohesion: Old (French) Wine in New Bottles? Urban Studies, 41 (7), pp. 1349–1365. FALUDI, A. (2005a): Territorial cohesion: an unidentified political objective. Town Planning Review, 76 (1), pp. 1–13. FALUDI, A. (2005b): Polycentric territorial cohesion policy. Town Planning Review, 76 (1), pp. 107–118. FALUDI, A. (2009a): A turning point in the development of European spatial planning? The ‘Territorial Agenda of the European Union’ and the ‘First Action Programme’. Progress in Planning, 71 (1), pp. 1–42. FALUDI, A. (2009b): Territorial Cohesion under the Looking Glass, http://ec.europa.eu/regional_policy/archive/consultation/terco/pdf/lookingglass.pdf (27/3/2014). FALUDI, A. (2010): Cohesion, coherence, cooperation: European spatial planning coming of age? Routledge, Oxon, 207 p.

169

FALUDI, A. (2013): Territorial cohesion, territorialism, territoriality, and soft planning: a critical review. Environment and Planning A, 45 (6), pp. 1302–1317. FALUDI, A., WATERHOUT, B. (2002): The Making of the European Spatial Development Perspective: No Masterplan! Routledge, London, 204 p. FARRUGIA, N., GALLINA, A. (2008): Developing Indicators of Territorial Cohesion. Federico Caffè Centre, Roskilde, 55 p. FIALA, P., PITROVÁ, M. (2010): Evropská unie. Centrum pro studium demokracie a kultury, Brno, 803 p. FINKA, M., KLUVÁNKOVÁ, T. (2015): Managing complexity of urban systems: A polycentric approach. Land Use Policy, 42, pp. 602–608. FINKA, M., KLUVÁNKOVÁ, T., ONDREJIČKA, V. (2015): Concept of Polycentric Governance for Fuzzy Soft Spaces as a Challenge for Central European Peripheral Spaces. In: LANG, T., HENN, S., SGIBNEV, W., EHRLICH, K. ed.: Understanding Geographies of Polarization and Peripheralization. Palgrave Macmillan, Hampshire, pp. 309–322. FOUCAULT, M. (1972): The Archaeology of Knowledge. Tavistock Publications Limited, UK, 245 p. FUJITA, M., KRUGMAN, P. (2004): The new economic geography: Past, present and the future. Papers in Regional Science, 83 (1), pp. 139–164. FUJITA, M., KRUGMAN, P., VENABLES, A. J. (1999): The Spatial Economy: Cities, Regions, and International Trade. MIT Press, Cambridge, MA, 381 p. GEDDES, P. (1915): Cities in Evolution. Williams and Margate, London, 446 p. GEORGE, S. (1996): Politics and Policy in the European Union. Oxford University, Oxford, 344 p. GLAESER, E. L. (2011): Triumph of the city. Penguin Press, New York, 352 p. GLØERSEN, E., BÖHME, K. (2011): Storylines on Territorial Cohesion, http://www.nordregio.se/en/Metameny/Nordregio-News-Thematic-Line/Europes-strive-for- Territorial-Cohesion/Storylines-on-Territorial-Cohesion/ (7/12/2015). GOEI, B. DE, BURGER, M. J., VAN OORT, F. G., KITSON, M. (2010): Functional polycentrism and urban network development in the Greater South East, United Kingdom: Evidence from commuting patterns, 1981–2001. Regional Studies, 44 (9), pp. 1149–1170. GREEN, N. (2007): Functional polycentricity: A formal definition in terms of social network analysis. Urban Studies, 44 (11), pp. 2077–2103. GROTH, N. B., SMIDT-JENSEN, S., NIELSEN, T. A. S. (2011): Polycentricity: An issue in local development strategies? Findings from the Baltic Sea region. European Planning Studies, 19 (5), pp. 727–751. HAGGET, P., CLIFF, A., FREY, A. (1977): Locational Analysis in Human Geography. Edward Arnold, London, 605 p.

170

HAGUE, C., KIRK, K. (2003): Polycentricity Scoping Study, http://www.northsearegion.eu/files/user/File/Events/Thematic%20Events/Polycentric%20Sem inar/ScopingPoly.pdf (6/2/2014). HAJER, M. (1993): Discourse coalitions and the institutionalisation of practice. The case of acid rain in Britain. In: FORESTER, J., FISCHER, F. ed.: The Argumentative Turn in Policy and Planning. Duke University Press, Durham, NC, pp. 43–76. HAJER, M. (1995): The politics of environmental discourse: Ecological modernization and the policy process. Oxford University Press, Oxford, 332 p. HAJER, M. (2000): Transnational networks as transnational policy discourse: Some observations on the politics of spatial development in Europe. In: SALET, W., FALUDI, A. ed.: The revival of strategic spatial planning. Royal Netherlands Academy of Arts and Sciences, Amsterdam, pp. 135–142. HALÁS, M., KLADIVO, P., ŠIMÁČEK, P., MINTÁLOVÁ, T. (2010): Delimitation of micro- regions in the Czech Republic by nodal relations. Moravian Geographical Reports, 18 (2), pp. 16–22. HALÁS, M., KLAPKA, P., TONEV, P., BEDNÁŘ, M. (2015): An alternative definition and use for the constraint function for rule-based methods of functional regionalisation. Environment and Planning A, 47 (5), pp. 1175–1191. HALÁSEK, D. (2012): Služby v obecném zájmu (Services of general interest). Acta Logistica, 2, pp. 23–28. HALL, P. (2002): Urban and Regional Planning. Routledge, London, 237 p. HALL, P., PAIN, K. (2006): The Polycentric Metropolis: learning from mega-city regions in Europe. Earthscan, London, 228 p. HAMEZ, G. (2005): Territorial Cohesion: How to Operationalize and Measure the Concept? Planning Theory & Practice, 6 (3), pp. 400–402. HAMILTON, I. (1976): Spatial structure in east European cities. In: FRENCH, R.A., HAMILTON, I. ed.: The socialist city. John Wiley & Sons, Chichester, pp. 195–262. HAMPL, M. (2005): Geografická organizace společnosti v České republice: Transformační procesy a jejich obecný kontext. Univerzita Karlova, Praha, 145 p. HANDY, S. (1992): Regional Versus Local Accessibility: Neo-Traditional Development and its Implications for Non-work Travel. Built Environment, 18 (4), pp. 253–267. HARVEY, D. (1982): The Limits to Capital. Blackwell, Oxford, 478 p. HAUGEN, K., HOLM, E., STRÖMGREN, M., VILHELMSON, B., WESTIN, K. (2012): Proximity, accessibility and choice: A matter of taste or condition? Papers in Regional Science, 91 (1), pp. 65–84. HÄZNERS, J., JIRGENA, H. (2013): Polycentricity measures and regional disparities. In: Annual 19th International Scientific Conference Proceedings: Research for Rural Development 2013, May 15–17. Latvia University of Agriculture, Jelgava.

171

HEALEY, P. (2009): The pragmatic tradition in planning thought. Journal of Planning Education and Research, 28 (3), pp. 277–292. HEFFNER, K., SOLGA, B. (2006): Small towns as local economic centres in rural areas. Bulletin of Geography (Socio-economic series), 6 (1), pp. 77–87. HEFFNER, K., TWARDZIK, M. ed. (2013): The Impact of Shopping Malls on the Outer Metropolitan Zones (The Example of the Silesian Voivodeship). Studia Regionalia, 37. Polish Academy of Sciences, Committee for Spatial Economy and Regional Planning, Warsaw. HINDERINK, J., TITUS, M. (2002): Small towns and regional development: Major findings and policy implications from comparative research. Urban Studies, 39 (3), pp. 379–391. HIRSCHMAN, A. (1958): The Strategy of Economic Development. Yale University Press, New Haven, 217 p. HOFFMAN, L. (1994): After the fall: crisis and renewal in urban planning in the Czech Republic. International Journal of Urban and Regional Research, 18 (4), pp. 691–702. HOLMES, J. H., HAGGETT, P. (1977): Graph Theory Interpretation of Flow Matrices: A Note on Maximization Procedures for Identifying Significant Links. Geographical Analysis, 9, pp. 388–399. HOOGHE, L. (1996): Cohesion Policy and European Integration: Building Multi-level Governance. Oxford University Press, New York, 458 p. HÜBNER, D. (2007): Territorial cohesion: Towards a clear and common understanding of the concept, http://europa.eu/rapid/pressReleasesAction.do?reference=SPEECH/07/743&format=HTML& aged=0&language=EN&guiLanguage=en (4/12/2015). HUDSON, R (2015): Uneven development, Socio-Spatial Polarization and Political Responses. In: LANG, T., HENN, S., SGIBNEV, W., EHRLICH, K. ed.: Understanding Geographies of Polarization and Peripheralization: Perspectives from Central and Eastern Europe and Beyond. Palgrave Macmillan, Hampshire, pp. 25–39. HUMER, A. (2014): Researching social services of general interest: An analytical framework derived from underlying policy systems. European Spatial Research and Policy, 21 (1), pp. 65–82. IDOS (2015): Jízdní řády, http://jizdnirady.idnes.cz/ (1/3/2015). IDS JMK (2015): South Moravian Integrated Public Transport System, http://www.idsjmk.cz/en/ (2/5/2015). ILLNER, M., ANDRLE, A. (1994): The Regional Aspect of Post-Communist Transformation in the Czech Republic. Czech Sociological Review, 2 (1), pp. 107–127. JENSES, O. B., RICHARDSON, T. (2001): Nested vision: new rationalities of space in European spatial planning. Regional Studies, 35 (8), pp. 703–717. KARLSSON, CH., OLSSON, M. (2006): The identification of functional regions: theory, methods, and applications. The Annals of Regional Science, 40 (1), pp. 1–18.

172

KEEBLE, D., OWENS, P. L., THOMPSON, C. (1982): Regional accessibility and economic potential in the European community. Regional Studies, 16 (6), pp. 419–432. KLAPKA, P., HALÁS, M., ERLEBACH, M., TONEV, P., BEDNÁŘ, M. (2014): A multistage agglomerative approach for defining functional regions of the Czech Republic: The use of 2001 commuting data. Moravian Geographical Reports, 22 (4), pp. 2–13. KLOOSTERMAN, R. C., LAMBREGTS, B. (2001): Clustering of economic activities in polycentric urban regions: The case of the Randstad. Urban Studies, 38 (4), pp. 717–732. KLOOSTERMAN, R. C., MUSTERD, S. (2001): The Polycentric Urban Region: Towards a Research Agenda. Urban Studies, 38 (4), pp. 623–633. KNOX, P. L., MAYER, H. (2012): Europe’s internal periphery: Small towns in the context of reflexive polycentricity. In: LORENTZEN, A., VAN HEUR, B. ed.: Cultural political economy of small cities. Routledge, London, pp. 142–157. KÖNIG, P. et al. (2009): Rozpočet a politiky Evropské unie: Příležitost pro změnu. C. H. Beck, Praha, 630 p. KOVÁCS, Z., TOSICS, I. (2014): Urban sprawl on the Danube: The impacts of suburbanization in Budapest. In: STANILOV, K., SÝKORA, L. ed.: Confronting suburbanization: Urban decentralization in postsocialist Central and Eastern Europe. Wiley Blackwell, Chichester, pp. 33–64. KRAFT, S., HALÁS, M., VANČURA, M. (2014): The delimitation of urban hinterlands based on transport flows: A case study of regional capitals in the Czech Republic. Moravian Geographical Reports, 22 (1), pp. 24–32. KRÄTKE, S. (2001): Strengthening the Polycentric Urban System in Europe: Conclusions from the ESDP. European Planning Studies, 9 (1), pp. 105–116. KRÄTKE, S. (2007): Metropolisation of the European economic territory as a consequence of increasing specialisation of urban agglomerations in the knowledge economy. European Planning Studies, 15 (1), pp. 1–27. KRUGMAN, P. (1991): Geography and Trade. MA: MIT Press, Cambridge, 142 p. KÜHN, M. (2015): Peripheralization: Theoretical Concepts Explaining Socio-Spatial Inequalities. European Planning Studies, 23 (2), pp. 367–378. KUNC, J., TONEV, P., FRANTÁL, B., SZCZYRBA, Z. (2012): Retail gravity models, shopping habits and shopping centres: The case of the Brno agglomeration (a contribution to the study of daily urban systems). Sociologický časopis, 48 (5), pp. 879–910. KUNZMANN, K. (2006): The Europeanization of Spatial Planning. In: ADAMS, N., ALDEN, J., HARRIS, N. ed.: Regional Development and Spatial Planning in an Enlarged European Union. Ashgate, Hampshire, pp. 43–63. KUNZMANN, K. R., WEGENER, M. (1991): The Pattern of Urbanisation in Western Europe 1960-90. Institut für Raumplanung, Universität Dortmund, Dortmund.

173

KUREK, S., WÓJTOWICZ, M., GAŁKA, J. (2015): The changing role of migration and natural increase in suburban population growth: The case of a non-capital post-socialist city (the Krakow Metropolitan Area, Poland). Moravian Geographical Reports, 23 (4), pp. 59–70. LAMBREGTS, B. (2009): The polycentric metropolis unpacked: Concepts, trends, and policy in the Randstad Holland, Doctoral dissertation, Amsterdam Institute for Metropolitan and International Development Studies, 208 p. LANG, T. (2015): Socio-economic and political responses to regional polarisation and socio- spatial peripheralisation in central and Eastern Europe: A research agenda. Hungarian Geographical Bulletin, 64 (3), pp. 171–185. LEFEBVRE, H. (1991): The Production of Space. Blackwell, Oxford, 454 p. LEVY, J. (1997): L’Europe, une géographie. Hachette, Paris. LÉVY, J. (2003): Aménagement du territoire, Capital spatial, Justice spatiale. In: LÉVY, J., LUSSAULT, M. ed.: Dictionnaire de la géographie et de l’espace des société. Belin, Paris. LIMTANAKOOL, N., SCHWANEN, T., DIJST, M. (2009): Developments in the Dutch Urban systems on the basis of flows. Regional Studies, 43 (2), pp. 179–196. LÓPEZ, E., GUTIÉRREZ, J., GÓMEZ, G. (2008): Measuring regional cohesion effects of large-scale transport infrastructure investments: An accessibility approach. European Planning Studies, 16 (2), pp. 277–301. LÖSCH, A. (1944): Die Räumliche Ordnung der Wirtschaft. Gustav Fischer, Jena, 380 p. LOUGHLIN, J. (1996): ‘Europe of the Regions’ and the Federalization of Europe. Publius, 26 (4), pp. 141–162. MAIER, K. (2004): Územní plánování. ČVUT, Praha, 85 p. MAIER, K., MULÍČEK, O., FRANKE, D. (2010): Vývoj regionalizace a vliv infrastruktur na atraktivitu území České republiky. Urbanismus a územní rozvoj, 8 (5), pp. 71–81. MALÍK, Z. (1976): Urbanizace a rozvoj systému osídlení. Výzkumný ústav výstavby a architektury, Praha, 110 p. MALÍK, Z., ČADA, R., LUKAŠÍK, M., MUSIL, J., ULIČNÝ, F., ZBOŘIL, M. (1968): Kritéria racionálního rozvoje osídlení. Výzkumný ústav výstavby a architektury, Brno. MALÍKOVÁ, L., KLOBOUČNÍK, M., BAČÍK, V., SPIŠIAK, P. (2015): Socio-economic changes in the borderlands of the Visegrad Group (V4) countries. Moravian Geographical Reports, 23 (2), pp. 26–37. MALÝ, J. (2016): Impact of Polycentric Urban Systems on Intra-regional Disparities: A Micro- regional Approach. European Planning Studies, 24 (1), pp. 116–138. MALÝ, J., MULÍČEK, O. (2016): European territorial cohesion policies: Parallels to socialist central planning? Moravian Geographical Reports, 24 (1), pp. 14–26.

174

MANZELLA, G. P., MENDEZ, C. (2009): The turning points of EU Cohesion policy, http://ec.europa.eu/regional_policy/archive/policy/future/pdf/8_manzella_final-formatted.pdf (21/8/2015). MAREŠ, J. (1988): Industrializace Československa – její klady a zápory. Sborník Československé geografické společnosti, 93 (3), pp. 183–198. MARKUSEN, A. (1999): Fuzzy concepts, scanty evidence, policy distance: The case for rigour and policy relevance in critical regional studies. Regional Studies, 33 (9), pp. 869–884. MARYÁŠ, J., KUNC, J., TONEV, P., SZCZYRBA, Z. (2014): Shopping and services related travel in the hinterland of Brno: changes from the socialist period to the present. Moravian Geographical Reports, 22 (3), pp. 18–28. MATOUŠEK, R. (2014): Prostorová spravedlnost. In: MATOUŠEK, R., OSMAN, R. ed.: Prostor(y) geografie. Karolinum, Praha, pp. 81–98. MAZA, A., VILLAVERDE, J. (2004): Regional disparities in the EU: Mobility and polarization. Applied Economics Letters, 11 (8), pp. 517–522. McGINNIS, M. D. ed. (1999): Polycentric governance and development: readings from the workshop in political theory and policy analysis. Univ. of Michigan Press, Ann Arbor, 448 p. MEDEIROS, E. (2011): Territorial Cohesion: a conceptual analysis, http://ww3.fl.ul.pt/pessoais/Eduardo_Medeiros/docs/PUB_PAP_EM_Territorial_Cohesion.pd f (20/1/2014). MEIJERS, E. (2007): From central place to network model: Theory and evidence of a paradigm change. Tijdschrift voor Economische en Sociale Geografie, 98 (2), pp. 245–259. MEIJERS, E. (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), pp. 2323–2342. MEIJERS, E. J., BURGER, M. J. (2015): Stretching the concept of ‘borrowed size’. Urban Studies, DOI: 10.1177/0042098015597642. MEIJERS, E. J., BURGER, M. J., HOOGERBRUGGE, M. M. (2016): Borrowing size in networks of cities: City size, network connectivity and metropolitan functions in Europe. Papers in Regional Science, 95 (1), pp. 181–198. MEIJERS, E., SANDBERG, K. (2006): Polycentric development to combat regional disparities? The relation between polycentricity and regional disparities in European countries. In: Proceedings of the 46th Congress of the European Regional Science Association. ERSA, Volos, pp. 1–20. MEIJERS, E., WATERHOUT, B., ZONNEVELD, W. (2005): Polycentric development policies in European countries: An introduction. Built Environment, 31 (2), pp. 97–102. MFČR (2013): Vyhláška č. 264/2013 Sb., http://www.mfcr.cz/cs/legislativa/legislativni- dokumenty/2013/vyhlaska-c-264-2013-sb-14254 (5/6/2014).

175

MMR (2013): Manuál služeb obecného hospodářského zájmu, http://www.uohs.cz/download/Sekce_VP/Zpravy/Manual-sluzeb-obecneho-hospodarskeho- zajmu.pdf (29/4/2015). MMR (2015a): Politika územního rozvoje České republiky, ve znění Aktualizace č. 1. Institute for Spatial Development, Brno. MMR (2015b): Vymezení ITI/IPRU metropolitních regionů v ČR. Internal document. Praha. MOLLE, W. (2007): European Cohesion Policy. Routledge, Oxon, 347 p. MULÍČEK, O. (2002): Suburbanizace v Brně a jeho okolí (Suburbanisation in Brno and its surroundings). In: SÝKORA, L. ed.: Suburbanizace a její sociální, ekonomické a ekologické důsledky (Suburbanization and Its Social, Economic, and Ecological Consequences). Ústav pro ekopolitiku, Praha, pp. 171–182. MULÍČEK, O., KOZEL, J. (2012): Metodika vymezení vztahově uzavřených funkčních regionů. CRR MU, Brno. MULÍČEK, O., SEIDENGLANZ, D., FRANKE, D., MALÝ, J. (2013): Vymezení funkčního území Brněnské metropolitní oblasti a Jihlavské sídelní aglomerace, https://www.brno.cz/iti/vymezeni-uzemi-brnenske-metropolitni-oblasti-bmo/ (5/5/2016). MULÍČEK, O., TOUŠEK, V. (2004): Changes of Brno Industry and their Urban Consequences. Bulletin of Geography, 3, pp. 61–70. MUSIL, J. (2001): Vývoj a plánování měst ve střední Evropě v období komunistických režimů – pohled historické sociologie. Sociologický časopis, 37 (3), pp. 275–296. MUSIL, J. (2002): Urbanizace českých zemí a socialismus. In: HORSKÁ, P., MAUR, E., MUSIL, J. ed.: Zrod velkoměsta: Urbanizace českých zemí a Evropa. Paseka, Praha-Litomyšl, pp. 237–331. MUSIL, J., LINK, J. (1976): Urbanizace v socialistických zemích ve světle mezinárodních srovnání. In: MUSIL, J. ed.: Otázky urbanizace. Sborník výzkumných prací kabinetu sociologie. Výzkumný ústav výstavby a architektury, Praha, pp. 9–30. NEDOVIC-BUDIC, Z. (2001): Adjustment of planning practice to the new eastern and central European context. Journal of the American Planning Association, 67 (1), pp. 38–52. NOZICK, R. (1974): Anarchy, State, and Utopia. Basic Books, New York, 384 p. NUGA, M., METSPALU, P., ORG, A., LEETMAA, K. (2015): Planning post-summurbia: From spontaneous pragmatism to collaborative planning? Moravian Geographical Reports, 23 (4), pp. 36–46. NYSTUEN, J. D., DACEY, M. F. (1961): A graph theory interpretation of nodal regions. Papers and Proceedings of the Regional Science Association, 7 (1), pp. 29–42. O'DOWD, L., ANDERSON, J., WILSON, T. M. (2013): New Borders for a Changing Europe: Cross-Border Cooperation and Governance. Routledge, London, 256 p. OECD (2008): Handbook on Constructing Composite Indicators: Methodology and User Guide. OECD, Paris, 158 p.

176

ØLYKKE, G. S., MØLLGAARD, P. (2013): What is a service of general economic interest? European Journal of Law and Economics, DOI: 10.1007/s10657-013-9426-8. OSMAN, R., FRANTÁL, B., KLUSÁČEK, P., KUNC, J., MARTINÁT, S. (2015): Factors affecting brownfield regeneration in post-socialist space: The case of the Czech Republic. Land Use Policy, 48, pp. 309–316. OUŘEDNÍČEK, M. (2003): Suburbanizace Prahy. Sociologický časopis, 39 (2), pp. 235–253. OUŘEDNÍČEK, M., ŠIMON, M., KOPEČNÁ, M. (2015): The reurbanisation concept and its utility for contemporary research on postsocialist cities: The case of the Czech Republic. Moravian Geographical Reports, 23 (4), pp. 26–35. OUŘEDNÍČEK, M., ŠPAČKOVÁ, P., FEŘTROVÁ, M. (2011): Changes in social milieu and quality of life in depopulating areas of the Czech Republic. Sociologický časopis, 47 (4), pp. 777–803. PALLA, V., MÁČEL, O., MALÍK, Z., LUKAŠÍK, M., ULIČNÝ, F., ČADA, R. (1962): Základní otázky osídlení v ČSSR. Výzkumný ústav výstavby a architektury, Brno. PARLIAMENT OF THE CZECH REPUBLIC (2009): Law no 222/2009, https://portal.gov.cz/app/zakony/?path=/portal/obcan/ (11/4/2015). PARR, J. B. (1987): Interaction in an urban system: Aspects of trade and commuting. Economic Geography, 63 (3), pp. 223–240. PARR, J. B. (2004): The Polycentric Urban Region: A Closer Inspection. Regional Studies, 38 (3), pp. 231–240. PARTRIDGE, M. D., RICKMAN, D. S., ALI, K., OLFERT, M. R. (2009): Do New Economic Geography agglomeration shadows underlie current population dynamics across the urban hierarchy? Papers in Regional Science, 88 (2), pp. 445–466. PERKMANN, M. (2003): Cross-Border Regions in Europe: Significance and Drivers of Regional Cross-Border Co-Operation. European Urban and Regional Studies, 10 (2), pp. 153– 171. PERKMANN, M. (2007): Policy Entrepreneurship and Multi-Level Governance: A Comparative Study of European Cross-Border Regions. Environment and Planning C: Government and Policy, 25 (6), pp. 861–879. PETERS, D. (2003): Cohesion, Polycentricity, Missing Links and Bottlenecks: Conflicting Spatial Storylines for Pan-European Transport Investments. European Planning Studies, 11 (3), pp. 317–339. PEYRONY, J. (2007): Territorial cohesion and the European model of society: French perspectives. In: FALUDI, A. ed.: Territorial Cohesion and the European Model of Society. Lincoln Institute of Land Policy, Cambridge, Massachusetts, pp. 61–79. PHELPS, N. A., FALLON, R. J., WILLIAMS, C. L. (2001): Small firms, borrowed size and the urban-rural shift. Regional Studies, 35 (7), pp. 613–624.

177

PORTNOV, B. A., FELSENSTEIN, D. (2005): Measures of regional inequality for small countries. In: FELSENSTEIN, D., PORTNOV, B. A. ed.: Regional Disparities in Small Countries. Springer, Heidelberg, pp. 47–62. PRED, A. (1977): City Systems in Advanced Economies: Past Growth, Present Processes and Future Development Options. Hutchinson, London, 256 p. PRESTON, J., RAJÉ, F. (2007): Accessibility, mobility and transport-related social exclusion. Journal of Transport Geography, 15 (3), pp. 151–160. RAFFINETTI, E., SILETTI, E., VERNIZZI, A. (2014): On the Gini coefficient normalization when attributes with negative values are considered. Statistical Methods and Applications, 24 (3), pp. 507–521. RAUHUT, D., LUDLOW, D. (2013): Services of General Interest and Territorial Cohesion: What, How and by Whom? Romanian Journal of Regional Science, 7, pp. 108–123. RAWLS, J. (1971): A Theory of Justice. Harvard University Press, London, 607 p. REPUS (2007): Regional Polycentric Urban System. Final report: Interreg III B Strategy for a Regional Polycentric Urban System in Central Eastern Europe Economic Integrating Zone. RePUS, 111 p. ROBERT, J. (2007): The Origins of Territorial Cohesion and the Vagaries of Its Trajectory. In: FALUDI, A. ed.: Territorial Cohesion and the European Model of Society. Lincoln Institute of Land Policy, Cambridge, Massachusetts, pp. 23–35. RODRIGUEZ, D. A., ROGERS, J. (2014): Can housing and accessibility information influence residential location choice and travel behavior? An experimental study. Environment and Planning B: Planning and Design, 41 (3), pp. 534–550. RYŠAVÝ, Z., DVOŘÁK, B., ŠIPLER, V. (1992): Posouzení dřívějších koncepcí územního rozvoje České republiky. Výzkumný ústav výstavby a architektury, Praha. SALONE, C. (2004): The role, specific situation and potentials of urban areas as nodes in a polycentric development. Paper presented at the national workshop “I partners italiani nel progetto ESPON”, Roma, 15 January 2004. SAPIR, A., AGHION, P., BERTOLA, G., HELLWIG, M., PISANI-FERRY, J., ROSATI, D., VIÑALS, J., WALLACE, H. (2004): An agenda for a growing Europe: The Sapir Report. Oxford University Press, Oxford, 234 p. SARMIENTO-MIRWALDT, K. (2015): Can multiple streams predict the territorial cohesion debate in the EU? European Urban and Regional Studies, 22 (4), pp. 431–445. SAUTER, W. (2008): Services of General Economic Interest and Universal Service in EU Law. European Law Review, TILEC Discussion Paper No. 2008-017. SCHÖN, P. (2005): Territorial Cohesion in Europe? Planning Theory & Practice, 6 (3), pp. 389–400. SCHWANEN, T. (2002): Urban form and commuting behaviour: a cross-European perspective. Tijdschrift voor economische en sociale geografie, 93 (3), pp. 336–343.

178

SEIDENGLANZ, D. (2010): Transport relations among settlement centres in the eastern part of the Czech Republic as a potential for polycentricity. Acta Universitatis Carolinae, Geographica, 45 (1), pp. 75–89. SERVILLO, L. (2010): Territorial Cohesion Discourses: Hegemonic Strategic Concepts in European Spatial Planning. Planning Theory & Practice, 11 (3), pp. 397–416. SERVILLO, L., ATKINSON, R., RUSSO, A. P. (2012): Territorial attractiveness in EU urban and spatial policy: A critical review and future research agenda. European Urban and Regional Studies, 19 (4), pp. 349–365. SHANKAR, R., SHAH, A. (2003): Bridging the economic divide within countries: A scorecard on the performance of regional policies in reducing regional income disparities. World Development, 31 (8), pp. 1421–1441. SKOKAN, K. (2008): Územní soudržnost v Evropě. Disputationes Scientificae Universitatis Catholicae in Ružomberok, 8 (1), pp. 85–95. SOJA, E. (1980): The socio-spatial dialectic. Annals of the Association of American Geographers, 70 (2), pp. 207–225. SOJA, E. (2009): The city and spatial justice. justice spatiale|spatial justice, 1, 5 p. SOJA, E. (2010): Seeking spatial justice. University Of Minnesota Press, Minneapolis, 221 p. SOKOL, M. (2001): Central and Eastern Europe a Decade After the Fall of State-socialism: Regional Dimensions of Transition Processes. Regional Studies, 35 (7), pp. 645–655. STEIN, S. M., HARPER, T. L. (2012): Creativity and innovation: Divergence and convergence in pragmatic dialogical planning. Journal of Planning Education and Research, 32 (1), pp. 5– 17. SÝKORA, L., BOUZAROVSKI, S. (2012): Multiple Transformations: Conceptualising the Post-communist Urban Transition. Urban Studies, 49 (1), pp. 43–60. SÝKORA, L., MULÍČEK, O. (2009): The micro-regional nature of functional urban areas (FUAs): lessons from the analysis of the Czech urban and regional system. Urban Research & Practice, 2, pp. 287–307. SÝKORA, L., MULÍČEK, O. (2012): Urbanizace a suburbanizace v Česku na počátku 21. století. Urbanismus a územní rozvoj, 15 (5), pp. 27–38. SÝKORA, L., MULÍČEK, O. (2014): Prague: Urban Growth and Regional Sprawl. In: STANILOV, K., SÝKORA, L. ed.: Confronting Suburbanization: Urban Decentralization in Postsocialist Central and Eastern Europe. Wiley Blackwell, Chichester, pp. 133–162. SÝKORA, L., MULÍČEK, O., MAIER, K. (2009): City regions and polycentric territorial development: concepts and practice. Urban Research & Practice, 2 (3), pp. 233–239. SÝKORA, L., OUŘEDNÍČEK, M. (2007): Sprawling Post-Communist Metropolis: commercial and residential suburbanisation in Prague and Brno, the Czech Republic. In: RAZIN, E., DIJST, M., VÁZQUEZ, C. ed.: Employment Deconcentration in European Metropolitan Areas. Springer, Dordrecht, pp. 209–233.

179

SZCZEPAŃSKI, J., FURDYNA, A. (1977): Early Stages of Socialist Industrialization and Changes in Social Class Structure. International Journal of Sociology, 7 (3/4), pp. 11–36. SZELENYI, I. (1996): Cities under socialism – and after. In: ANDRUSZ, G., HARLOE, M., SZELENYI, I. ed.: Cities after socialism. Urban and regional change and conflict in post- socialist societies. Willey-Blackwell, Oxford, pp. 286–336. TAMMARU, T. (2001): Suburban Growth and Suburbanisation under Central Planning: The Case of Soviet Estonia. Urban Studies, 38 (8), pp. 1341–1357. TAYLOR, P. J., EVANS, D. M., PAIN, K. (2008): Application of the interlocking network model to megacity regions: Measuring polycentricity within and beyond city-regions. Regional Studies, 42 (8), pp. 1079–1093. TERPLAN (1981): Koncepce sídelního skeletu: Výchozí podklady. Praha. TERPLAN (1985): Brněnská sídelní regionální aglomerace. Územní plán VÚC. Praha. TERVO, H. (2010): Cities, hinterlands and agglomeration shadows: Spatial developments in Finland during 1880-2004. Explorations in Economic History, 47 (4), pp. 476–486. THE SOUTH MORAVIAN REGION (2012): Komplexní dotazníkové šetření v obcích Jihomoravského kraje k 31. 12. 2012, http://www.kr- jihomoravsky.cz/Default.aspx?ID=201445&TypeID=2 (2/5/2015). TUROK, I., BAILEY, N. (2004): The theory of polynuclear urban regions and its application to central Scotland. European Planning Studies, 12 (3), pp. 371–389. UAD STUDIO (2014): Územní studie sídelní struktury Jihomoravského kraje. Brno. USB (2015): Zásady územního rozvoje Jihomoravského kraje. In preparation. Brno. VAISHAR, A., ŠŤASTNÁ, M., STONAWSKÁ, K. (2015): Small towns - Engines of rural development in the South-Moravian region (Czechia): An analysis of the demographic development. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, 63 (4), pp. 1395–1405. VAISHAR, A., ZAPLETALOVÁ, J. (2009): Small towns as centres of rural micro-regions. European Countryside, 1 (2), pp. 70–81. VAN DER KNAAP, G. A. (2007): Urban network formation under conditions of uncertainty. In: TAYLOR, P. J., DERUDDER, B., SAEY, P., WITLOX, F. ed.: Cities in Globalization: Practices, Policies and Theories. Routledge, London, pp. 84–101. VAN DER LAAN, L. (1998): Changing urban systems: An empirical analysis at two spatial levels. Regional Studies, 32 (3), pp. 235–247. VAN DER LAAN, L., SCHALKE, R. (2001): Reality versus Policy: The Delineation and Testing of Local Labour Market and Spatial Policy Areas. European Planning Studies, 9 (2), pp. 201–221. VAN DER MEER, L. (1998): Red octopus. In: BLAAS, W. ed.: A new perspective for European spatial development policies. Ashgate, Aldershot, pp. 9–19.

180

VAN LEEUWEN, E. S., RIETVELD, P. (2011): Spatial consumer behaviour in small and medium-sized towns. Regional Studies, 45 (8), pp. 1107–1119. VAN NUFFEL, N. (2007): Determination of the Number of Significant Flows in Origin– Destination Specific Analysis: The Case of Commuting in Flanders. Regional Studies, 41 (4), pp. 509–524. VAN NUFFEL, N., DERUDDER, B., WITLOX, F. (2009): Even important connections are not always meaningful: on the use of a polarisation measure in a typology of European cities in air transport networks. Tijdschrift voor Economische en Sociale Geografie, 101 (3), pp. 338– 348. VANOLO, A. (2010): European Spatial Planning Between Competitiveness and Territorial Cohesion: Shadows of Neo-liberalism. European Planning Studies, 18 (8), pp. 1301–1315. VASANEN, A. (2013): Spatial integration and functional balance in polycentric urban systems: A multi-scalar approach. Tijdschrift voor Economische en Sociale Geografie, 104 (4), pp. 410– 425. VENERI, P., BURGALASSI, D. (2012): Questioning polycentric development and its effects: Issues of definition and measurement for the Italian NUTS 2 Regions. European Planning Studies, 20 (6), pp. 1017–1037. VICKERMAN, R., SPIEKERMANN, K., WEGENER, M. (1999): Accessibility and Economic Development in Europe. Regional Studies, 33 (1), pp. 1–15. VÚVA (1979): Zásady a pravidla územního plánování. Výzkumný ústav výstavby a architektury, Urbanistické pracoviště Brno. WANG, F., TANG, Q. (2013): Planning toward equal accessibility to services: A quadratic programming approach. Environment and Planning B: Planning and Design, 40 (2), pp. 195– 212. WATERHOUT, B. (2002): Polycentric development: What is behind it? In: FALUDI, A. ed.: European Spatial Planning. Lincoln Institute of Land Policy, Cambridge, MA, pp. 83–103. WATERHOUT, B. (2007): Territorial cohesion: the underlying discourses. In: FALUDI, A. ed.: Territorial Cohesion and the European Model of Society. Lincoln Institute of Land Policy, Cambridge, MA, pp. 37–59. WATERHOUT, B. (2008): The institutionalisation of European spatial planning. IOS Press, Amsterdam, 225 p. WEBBER, D., WHITE, P. (2003): Regional factor price convergence across four major European countries. Regional Studies, 37 (8), pp. 773–782. WITTEN, K., EXETER, D., FIELD, A. (2003): The Quality of Urban Environments: Mapping Variation in Access to Community Resources. Urban Studies, 40 (1), pp. 161–177. WU, F. (2003): Transitional Cities. Commentary. Environment and Planning A, 35, pp. 1331– 1338.

181

ZEKERI, A. A. (1994): Adoption of economic development strategies in small towns and rural areas: Effects of past community action. Journal of Rural Studies, 10 (2), pp. 185–195. ZONNEVELD, W. (2000): Discursive aspects of strategic planning: A deconstruction of the ‘balanced competitiveness’ concept in European spatial planning. In: SALET, W., FALUDI, A. ed.: The revival of strategic spatial planning. Royal Netherlands Academy of Arts and Sciences, Amsterdam, pp. 267–280. ZONNEVELD, W., MEIJERS, E., WATERHOUT, B. ed. (2005): Polycentric development policies across Europe. Built Environment, 31 (2), pp. 93–173. ZONNEVELD, W., WATERHOUT, B. (2005): Visions on territorial cohesion. Town Planning Review, 76 (1), pp. 15–27.

182

FIGURES

Fig. 1. The tree model used to define the multidimensional phenomenon according to Abrahams (2014) and OECD (2008)...... 34 Fig. 2. The TEQUILA model of territorial cohesion...... 37 Fig. 3. The INTERCO model of territorial cohesion...... 38

Paper 1 Fig. 1. Geographical location of the South Moravian Region...... 67 Fig. 2. Comparison of urban regional agglomerations (URA)a delimited in 1976 and ITI/IPRU metropolitan regions (MR)b delimited in 2014...... 68 Fig. 3. Delimitation of settlement centres in the South Moravian Region by UAD Studio in 2014 and Terplan in 1985...... 72

Paper 2 Fig. 1. Urban system of the Czech Republic – municipalities with 10,000 and more inhabitants...... 83 Fig. 2. Change of the number of integrated municipalities between 1991 and 2011, the case of centres of significant flows in the Czech Republic...... 89 Fig. 3. Change of the total number of in-commuters between 1991 and 2011, the case of centres of significant flows in the Czech Republic...... 91

Paper 3 Fig. 1. The Czech functional regions based on work commuting flows in 2011...... 103 Fig. 2. The levels of morphological and functional polycentricity in the Czech functional regions...... 108 Fig. 3. Examples of positive and negative correlations between levels of morphological polycentricity and the specific indicator of intra-regional disparities measured by the Gini coefficient (G) and standard deviation (SD)...... 110

Paper 4 Fig. 1. The urban system of the South Moravian Region...... 121 Fig. 2. Transportation accessibility of selected services – the nearest facility of a particular service for each municipality in the South Moravian Region measured by the car time distance (left) and the number of public transport connections during a week-day (right) and sorted from a municipality with the highest accessibility (1) to a municipality with the lowest accessibility (672)...... 123 Fig. 3. The index of service function centrality for municipalities in the South Moravian Region...... 126 Fig. 4. Relative values of the index of service function centrality for small centres in the South Moravian Region...... 130

183

Paper 5 Fig. 1. The South Moravian Region...... 139 Fig. 2. The SGI accessibility TSGI; municipalities in the South Moravian Region...... 147

184

TABLES

Tab. 1. Definitions of territorial cohesion based on five ESPON projects and Territorial Agendas...... 35 Tab. 2. Storylines of territorial cohesion...... 44

Paper 1 Tab. 1. Categories of centres based on their function and regional significance (in 1985) and their presence in the South Moravian Region (SMR)...... 71 Tab. 2. Categories of centres based on production and service potential (in 2014) and their presence in the South Moravian Region (SMR)...... 71

Paper 2 Tab. 1. Ideal configurations of outgoing flows...... 84 Tab. 2. Work commuting significant flows in 1991, 2001 and 2011 in the Czech Republic. . 86 Tab. 3. Centres of significant flows in the Czech Republic, the change between 1991 and 2011...... 87 Tab. 4. Work commuting characteristics according to the size categories of centres, the case of the Czech Republic...... 88

Paper 3 Tab. 1. Size of the Czech functional regions and Pearson correlation coefficient between population size of the regions and levels of morphological (MP) and functional (FP) polycentricity...... 108 Tab. 2. Pearson correlation coefficient between polycentricity levels and disparity measures...... 113

Paper 4 Tab. 1. Distribution of services among the municipalities in the South Moravian Region. .. 122 Tab. 2. Distribution of services among the municipalities in the South Moravian Region. .. 125 Tab. 3. The index of service function centrality for small centres in the South Moravian Region...... 127 Tab. 4. Logistic regression model for the factors (independent variables) affected service function centrality of small centres in the South Moravian Region and BMA...... 128

Paper 5 Tab. 1. The model values of the number of facilities F, the score of centrality SC and the municipal centrality MC for services a, b, c and municipalities i and j...... 142 Tab. 2. The methodology used for calculating each type of distance d between the municipality and the centre of a particular service...... 143

185

Tab. 3. The score of centrality SC for a particular SGI; municipalities in the South Moravian Region...... 145 Tab. 4. The Pearson correlation coefficient between the accessibility of services based on SC category (SC 1 = 100<; SC 2 = 15-99; SC 3 = 0-14) and the development indicators; municipalities in the South Moravian Region...... 149 Tab. 5. Multiple regression analysis of the attractiveness of places (population growth, crude net migration rate and completed dwellings) and socio-economic conditions; municipalities in the South Moravian Region...... 150

186

LIST OF ABBREVIATIONS

AER Assembly of European Regions ATM Automated Teller Machine BMA Brno Metropolitan Area CAP Common Agricultural Policy CEMAT Conference Européenne de Ministers Responsable pour l’Aménagement du Territoire CF Cohesion Fund CoR Committee of the Regions CPMR Conference of Peripheral and Maritime Regions of Europe DATAR Délégation à l´Aménagement du Territoire et à l´Action Régionale DG REGIO Directorate-General for Regional and Urban Policy DG XVI Directorate-General for Regional Policies and Cohesion EAEC European Atomic Energy Community EAGGF European Agricultural Guidance and Guarantee Fund EC European Community ECSC European Coal and Steel Community EEC European Economic Community EESC European Economic and Social Committee EIB European Investment Bank ERDF European Regional Development Fund ESDP European Spatial Development Perspective ESF European Social Fund ESPON European Spatial Planning Observation Network/European Observation Network for Territorial Development and Cohesion ETCI European Territorial Cohesion Index EU European Union GDP Gross Domestic Product INTERCO Indicators of Territorial Cohesion IPRU Integrated Plans of Territorial Development ITI Integrated Territorial Investments LAU Local Administrative Unit

187

MLA Multiple Linkage Analysis MR Metropolitan Region NUTS Nomenclature des Unités Territoriales Statistiques OECD/JRC Organisation for Economic Co-operation and Development/Joint Research Centre PUR Polycentric Urban Region REGI Committee on Regional Development SEA Single European Act SGI Services of General Interest SGEI Services of General Economic Interest SMR South Moravian Region SSGI Social Services of General Interest TEQUILA Territorial Efficiency Quality Identity Layered Assessment TIA Territorial Impact Assessment TTWA Travel-to-Work Areas URA Urban Regional Agglomeration

188

APPENDICES

Appendix 1. Competitiveness-Oriented Scenario of the EU territory – Final image 2030.

Source: ESPON (2007)

189

Appendix 2. Cohesion-Oriented Scenario of the EU territory.

Source: ESPON (2007)

190

Appendix 3. Cartographic image of an open and polycentric Europe.

191

Appendix 3. Legend.

Source: ESPON (2014b)

192

Appendix 4. Contribution of authors to the presented papers.

Paper 1 MALÝ, J., MULÍČEK, O. (2016): European Territorial Cohesion Policies: Parallels to Socialist Central Planning? Moravian Geographical Reports, 24 (1), pp. 14–26. Jiří Malý and Ondřej Mulíček designed the study conception and the introduction. Jiří Malý and Ondřej Mulíček contributed to the theoretical framework. Both authors processed and analysed the data. Jiří Malý summarized the methodology and wrote the conclusion. Ondřej Mulíček provided comments on the manuscript.

……………………….. ……………………….. Jiří Malý (60%) Ondřej Mulíček (40%)

Paper2 MALÝ, J., MULÍČEK, O.: manuscript (under review in European Urban and Regional Studies) Both authors participated on the study conception and the theoretical background. Ondřej Mulíček processed the data and Jiří Malý performed the analysis. Jiří Malý described and interpreted the results. Ondřej Mulíček came up with the main concluding remarks and provided comments on the final version of the paper.

……………………….. ……………………….. Jiří Malý (60%) Ondřej Mulíček (40%)

Paper3 MALÝ, J. (2016): Impact of Polycentric Urban Systems on Intra-regional Disparities: A Micro- regional Approach. European Planning Studies, 24 (1), pp. 116–138. Jiří Malý conceived the idea, carried out the analysis and wrote the manuscript.

……………………….. Jiří Malý

Paper 4 MALÝ, J. (2016): Small towns in the context of ‘borrowed size’ and ‘agglomeration shadow’ debates: The case of the South Moravian Region (Czech Republic). European Countryside, in press. Jiří Malý conceived the idea, carried out the analysis and wrote the manuscript.

……………………….. Jiří Malý

193

Paper 5 MALÝ, J. (2016): Questioning territorial cohesion: (Un)equal access to services of general interest. Papers in Regional Science, DOI: 10.1111/pirs.12250. Jiří Malý conceived the idea, carried out the analysis and wrote the manuscript.

……………………….. Jiří Malý

194