CITIES AND REGIONS IN COMPETITION? NEGOTIATIONS FOR THE 2014-2020 PROGRAMMING PERIOD IN THE CZECH REPUBLIC1⃰

Abstract Over the past two decades, the concept of multi-level governance (MLG) has been increasingly discussed by scholars in the fi eld of European integration. While Gary Marks wrote about a four-level arrangement (supranational, national, regional, and local), over time the regional and local levels often became lumped together as ‘substate actors’ and so easily conceptually interchangeable. This text, however, shows the fallaciousness of this reasoning. In certain circumstances we can fi nd a competitive relationship between cities and regions, positioning themselves against each other for resources and access to national and supranational fora, especially in the context of the new regionalism. The cities have been given substantial support from the European Commission in recent years and we argue that this new constellation may have a remarkable infl uence on relations and possibly also lead to confl icts among local and regional actors in EU multi-level governance. This was possible to be clearly seen in the Czech Republic (CR) between 2012 and 2014, when heated negotiations took place regarding the implementation of the Integrated Territorial Investment (ITI), a fi nancial instrument of EU Cohesion Policy which was implemented on the substate level – i.e., in cities and regions. In the CR this competition took place in a specifi c context, which also infl uenced its outcome.


Introduction
One certain fl aw or problem of the concept of multi-level governance (MLG) is that as the use of the concept increases, the formulation of diff erent expressions and defi nitions increases accordingly.This situation can be observed even in defi ning the level on which governance occurs.While Gary Marks (1993, p. 392) spoke at the outset about four levels of governance -European, national, regional, and local -subsequent authors often connected the two latt er levels under the collective term 'subnational' or 'substate', which has been very misleading at the least.The use of the expression 'substate' can, in some specifi c cases, create the impression that there exists a coherent group of actors (comp.Callanan and Tatham, 2014) with more or less similar interests and behavior, which is not necessarily the case.This 'substate' term derives from the three-level concept of policymaking.Kassim (2005) uses this approach when discussing the Europeanization of substate authorities, which he defi nes as subnational states, regions, and municipalities (similarly to Jeff ery, 2000).The fact that local arena is described relatively haphazardly in defi nitions of MLG is rather alarming, as it obscures the clarity of the concept.The answer why this happens exists.As research on Czech and German cities shows (Havlík, 2013), it is not possible to speak of 'typical' actors in MLG in the case of cities, because the nature of their actions is quite heterogeneous.This statement was valid around 2010, and is more or less valid still today, even though the role of major cities (with metropolitan suburbs) has changed since that time.The role of major cities (typically those with more than about 100,000 inhabitants) as clearly emancipated actors emerged decisively after 2007 (and more strongly after 2014) based on the new period of the EU regional policy, which following the Lisbon strategy provided them with much stronger tools in implementing the EU structural funds (the so-called 'territorial dimension', see below).The idea behind this was to raise competitiveness across the EU by making 'new internationally signifi cant territories of economic dynamism' (Deas and Lord, 2006, pp.1848-1849).This empowering of cities, above all the empowering of the so-called 'urban regions' or 'metropolitan' areas, has been often discussed in previous literature on 'new regionalism' (Harrison, 2007; Deas and Lord, 2006; Zimmermann and Heinelt, 2012; Zimmermann, 2008; Bafarasat, 2016, etc.).This text's ambition is therefore to off er a new perspective on the relations between substate actors; more specifi cally, the text shows that in certain circumstances we can fi nd a competitive relationship between cities and regions.It was possible to clearly see this in the Czech Republic between 2012 and 2014, when heated negotiations took place regarding the implementation of the Integrated Territorial Investment (ITI), a fi nancial instrument of EU Cohesion Policy which is implemented on the substate level -i.e., in cities, metropolitan areas, and regions.This new instrument is interesting for a variety of reasons (see below), but particularly because it confi rms the above-mentioned att empts of the European Commission to broaden the implementation of cohesion policy, implementing it not only region-ally, but also to metropolitan (subregional and local) areas (comp.Zimmermann and  Heinelt, 2012).This presents a challenge to current understandings of the concept of multi-level governance from the point of view of the division of individual levels of governance.In other words, it is necessary to address possible changes in the roles of substate actors, which were brought about in the context of the new forms of EU Regional Policy during the 2014-2020 programming period.Speaking about metropolitan governance is of course 'nothing new', but it may be worthwhile to update this discussion, since cities have been given substantial support from the European Commission in recent years.We argue that this new constellation may have a remarkable infl uence on relations and possibly also lead to confl icts among local and regional actors in EU multi-level governance.
In light of this, the Czech Republic is a prime case for study.First, as one of the new Member States of the EU, it is a major recipient of fi nancial support under the cohesion policy.It is precisely due to the fact that substate actors have a tangible chance to acquire a substantial amount of money from the structural funds that they have more than suffi cient reasons to fi ght for this money.The structure of the Czech state also makes it a good fi t for research.As a unitary state, the regional level did not play a signifi cant role before entry into the EU, and therefore regions had a similar initial position to cities during the EU accession.As a part of the post-communist heritage, cities lacked a history of any real autonomy and the regions were newly formed in the 1990s after a decade of internal political disputes about their formation, number, and competencies.Both cities and regions had to adapt to specifi c mechanisms of both direct and indirect communication with EU actors and interest promotion bodies.Moreover, this adaptation has been further complicated by problems of a heritage of political and administrative culture shared by various post-communist 'new' EU member states.The interaction of all these circumstances, combined with the amount of European money that cities and regions compete over, make the Czech Republic an interesting object of research.
In the next parts of the text, we are going to present step by step the features of Czech substate governance and the relations between regions and cities.In the second section, we put our case study in the context of the new regionalism theory and show that the changes of the substate landscape and of the relations between cities and regions are undoubtedly driven by the activities of the EU Commission.In the third section, we present the specifi cs of the Czech cohesion policy and discuss the infl uence of these specifi cs on the performance of Czech substate actors in the EU multi-level governance.Then, after explaining our methodology we present in the fi fth section our fi ndings.First of all, we present the discussion about the implementation of the ITI fi nancial tool, than we present and explain the ways which cities and regions used for their interest articulation.Finally, we show that the ITI tool was a crucial and divisive factor for the relations of cities and regions in the context of particular problematic factors concerning the Czech Republic.

Subnational actors and their strategies of interest articulation in the context of MLG
Subnational actors have a wide range of options in the process of interest articulation in the context of multi-level governance of the EU.For greater clarity, these options can be positioned within various existing concepts (MLG, Europeanization, paradiplomacy, the partnership principle, etc.) which, rather than identifying the points that cities and regions can use, off er a basis to understand the activities of these actors.In this text, however, it functions as a departure point for formulating the argument that cities may be seen as being on equal footing with and, in some cases, even competitors to regions.
Activities of substate actors can be further divided as domestically-and foreign-oriented on the basis of existing empirical studies (comp.Münch, 2006, etc.).In the case of domestic activities, this is above all interaction with the relevant government ministry, either directly on a bilateral basis, or more commonly in the form of membership in an organization representing the interests of such actors (association of regions, organizations of towns and cities).In their 'paradiplomacy' (Soldatos  and Michelmann, 1992), substate actors see increasing regionalization as a 'throwing down of the gauntlet' to increase their activities and att empts to infl uence the content of European legislation.In this process, various possible entry points exist; just as in the case of interest articulation on the state level, these can be approached bilaterally (direct contact to EU institutions, participation in public consultations, etc.) as well as multilaterally (membership in the Committ ee of Regions, CEMR, Eurocities, etc.).
The term 'strategy', used in the title of this section, is imprecise and often distorted.As previous research (Havlík, 2013) shows, while some substate actors have a detailed and elaborate approach to 'European' interest advocacy, 1 others hardly have any 'strategy' at all.In this context, it is interesting that the main line dividing successful from unsuccessful substate actors is neither the structure of the executive branch in a given state (unitary vs. federal), nor the 'dichotomy' between old and new EU Member States, nor higher or lower levels of corporatism in a given state.Often, success of a specifi c substate actor is determined by how proactive it is in its approach, and by its own decision to actively campaign for its interests regarding individual European policies (comp.Callanan and Tatham, 2014).
This fi nding is important for further argumentation.It is clear that while initial conditions for individual substate actors vary (some are larger, wealthier, and undoubtedly more infl uential), each of these actors have at least a few options to articulate their interests.Moreover, MLG rejects presenting cities as actors positioned vertically below regions, but instead places them in a circle beside regions; this reveals their options in a completely diff erent light than they would be in a more 'conserva-tive', linear understanding.However, cities did not always have the same opportunities in MLG as regions.Their placement on the same level as regions in MLG is accurate, but, from a quantitative point of view, they were in charge of a smaller amount of money in recent decades.The concept of 'new internationally signifi cant territories of economic dynamism' (Deas and Lord, 2006, pp.1848-1849), as mentioned above in speaking about the new regionalism, changed their opportunities and possibly their position in MLG as well.Although the European Commission does not give the cities the right to implement the ITI directly, they do have the opportunity to enter into competition (with other regional or subregional actors) with a serious chance to win.In other words, the EU Commission's latest trend in making regional policy has been to establish incentive structures for regional and subregional actors in order to let the best win, but without directly saying who the best is and who exactly the implementing authority should be.
In this text, we argue that as long as cities did not get any real opportunity to decide about the implementation of a substantial part of structural funds on their own, their participation in MLG seemed rather formal and fi t for rather theoretical disputations of political scientists.But in the context of the gradually developing new regionalism, we may speak during the last decade about cities as 'true actors' of MLG, as they have been strongly supported by the European Commission (comp.Zimmermann, 2008).We further argue that this granting of support may give rise to tensions between regions and cities, which makes further research of their relationship in this fi eld welcomed (comp.Deas and Lord, 2006, p. 1862).
Since 2006, when Deas and Lord wrote that there is 'litt le or no evidence of significant confl ict' of 'new regions' with 'old regions', the respective literature has become more profound (Deas and Lord, 2006; Zimmermann and Heinelt, 2012; Bafarasat, 2016).In spite of this, case studies about the confl ict of new and old regions in the EU money distribution process are still lacking.The goal of a specifi c case study in the context of the Czech Republic is thus to examine the relationships between regions and cities in detail, particularly in a country in which regions as actors have not been historically more infl uential than cities.This makes it an ideal and interesting area for research.

Development of cohesion policy in the Czech Republic
For Czechoslovakia and later for the Czech Republic (CR), the year 1989 signifi ed a historic break not only in the sense of the end of single-party rule and the transition to democracy, but also slowly facilitated the debate on the style of governance.Regionalization and the development of regional policy were not excluded from this debate, addressing the goal of reducing regional disparities.However, in the fi rst half of the 1990s, this newly formed goal was not visibly achieved.To a large extent, actual work toward it was connected to the Czech Republic's drawing nearer to the EU in the second half of the 1990s, as a functioning regional administrative organ was one of the conditions for entry (Dočkal and Kozlová, 2006).Thus in the CR, regionalization often has correlated with increasing Europeanization.On 1 January 1996, the Ministry of Regional Development was established, and in the following year a constitutional amendment was passed to create and implement 14 administrative units known as 'kraje', in the year 2000.A similar regional division of the country had existed in the communist era as well, but in light of the centralized character of state administration, the kraje had rather formal powers and competences than a real power.In addition to this, the new structure and number of regions does not copy the structure from the communist era: some new regions were established, and some of the previous regions were divided into smaller units.This means that both regions which had some tradition of existence as 'kraje' even during the communist era and some regions, which were newly established in the 1990's, exist.Also it quickly became clear that the att empts of the CR to create a situation in which the kraje would be recognized as NUTS II regions would be generally unsuccessful, 2 and in an eff ort to secure money from the structural funds, some of the kraje were merged for NUTS II purposes; this resulted in units that were even more artifi cial, lacking any homogenous identity or historical borders (comp.Heinelt and  Lang, 2011; Dočkal and Kozlová, 2006; Benedek and Bajtalan, 2015).This is not to say that the kraje at the time of their creation had a natural position or authority -they did not.The law on kraje (Zákon o krajích č.129/2000) indeed gave them relatively limited powers; with the exception of the right to propose laws in the Parliament, these were certain limited powers in transportation, education, and health-care. 3ities had in that time certain amount of experience with interest articulation, which they had gained at the beginning of the 20 th century (in Austria-Hungary) when they began to form associations.In the modern era, cities renewed their associative actions shortly after the Velvet Revolution as the 'Union of Towns and Municipalities of the CR' (Svaz měst a obcí ČR, UTMCR) and shortly thereafter were seen as a traditional partner of the government.The kraje had to form a similar structure from scratch.The Association of Regions of the Czech Republic (Asociace krajů ČR) became a respected and active partner of the government relatively quickly, in part due to the low number of kraje; the association thus was able to more easily aggregate the positions of its members than the UTMCR.
But even at the most general level, it cannot be stated that the kraje were a priori more infl uential actors, more likely to succeed in their att empts to lobby for their interests than cities were.In particular, large statutory cities did not face the same obstacles as smaller cities -above all, a lack of population, which implies an unatt ractive position for industrial development and employment and a generally lower 'pull' in negotiations with central organs of the state.Brno and Ostrava, the second and third largest cities in the Czech Republic, certainly did not experience these obstacles and to a large extent represented equal partners to their regional counterparts. 4mong other things, the position of kraje and cities is connected to how the EU structural funds were implemented in the CR.The means by which the Czech Republic set up its operational programs for securing of monies from the structural funds has always been a subject of discussion, particularly concerning the degree of decentralization of the implementation structure.The country's fi rst period of regional policy was brief, from the date of its entry into the EU (May 2004) to 2006.Individual NUTS II regions did not have access to their own operational programs, but drew funds from a common program, the Joint Regional Operational Program (Společný regionální operační program, JROP).Conversely, during 2007-2013 seven regional operational programs (ROPs) appeared in addition to the nationwide thematic operational programs, indicative of the clear decentralization of the entire process of European subsidy acquisition.However, the process of implementation of European regional policy was at this time accompanied by widely publicized corruption scandals that embroiled many NUTS II regions.Although these cases of corruption were offi cially linked to some NUTS II regions, there always have been connotations to the politicians of kraje, since the kraje are a constitutive part of NUTS II regions and in some cases (see above), a kraj has the same territory as the NUTS II region.Partly in response to these scandals, regional programs for the 2014-2020 period were canceled and the country shifted back to centralization of the whole process under the JROP heading.
However, the kraje 'lost' in their batt le over the implementation of another powerful tool.So-called 'Integrated Territorial Investment', which (like the original ROP) was conceived of as a similar decentralized instrument, was awarded to major cities after a protracted fi ght.Precisely this process of the rise of competition between kraje and big cities is the topic of the following portion of the text.

Research design
In connection with the theoretical and empirical fi ndings described above, this article att empts to answer the following research questions: 1. Which strategies, approaches, and mechanisms did the given substate actors (kraje and relevant cities 5 ) use to advocate for their interests in the area of im-plementation of the EU's regional policy tools in the case of negotiations on ITI implementation, and how does this occur on the domestic as well as on the international/EU levels?And why? 2. What were the reasons for the confl ict between the old and new regions (kraje and major cities) in the Czech Republic and which main factors infl uenced the outcome of the confl ict?
These research questions help us to discover specifi c diff erences in the tactics of the relevant cities in the CR on the one hand, and kraje on the other, and at the same time help in directing att ention to important factors which infl uence these diff erences in approach, both in the case of cities as well as in the case of kraje.
First off , regarding cities, it is logical to study the cities of the seven largest metropolitan areas in the CR as actors qualifi ed to use the ITI mechanism (districts of Ústí-Chomutov, Plzeň, Prague, Hradec Králové-Pardubice, Brno, Olomouc, and Ostrava).Of these, the main cities of fi ve of the districts agreed to participate in the research project (Prague, Plzeň, Brno, Olomouc, and Ostrava). 6This group of major cities in their corresponding metropolitan areas represents a suffi ciently representative sample.In the case of the kraje, out of a total of 14, fi ve took part in the survey (more than one-third, and thus they represent an appropriate sample): The Capital City of Prague, Moravia-Silesia, and South Moravia are three of the top four most-populated kraje in the CR (each with over 1 million residents), while the Hradec Králové Region and the Karlovarský (Carlsbad) Region are less populated (the Karlovarský Region is the least populated of all 14).The sample demonstrates the diverse character of the individual kraje, not only from a demographic, but also from a geographic point of view.Moreover, the sample takes into account varying structures of population distribution.For example, South Moravia has one dominant city (Brno) in terms of population and economic potential, which makes it a magnet for commuters, etc.In contrast, in Moravia-Silesia there is a larger number of cities that have grown into a large agglomeration.Because such characteristics may have a signifi cant impact on methods of interest articulation, they are critical to the selection of an appropriate sample.One specifi c case included in the research is the Capital City of Prague, which is at once a major city and a de jure kraj, as well as the capital.This gives it a unique position in terms of socio-economic characteristics, as well as in bargaining processes, both in national and in the EU-level structures.
Methodologically, this research employs semi-structured qualitative interviews, or more precisely 'expert interviews' (Litt ig, 2009).This technique was chosen in light of the goals of the research, which were focused on the practical utilization of existing mechanisms at the national as well as at the European Union level, on what specifi c negotiation and communication practices occur through a given channel, and what factors infl uence these approaches. 7Interviews were undertaken with employees of city halls as well as offi ces of the kraje. 8Emphasis was placed on the selection of employees in departments which were directly concerned with regional or local development, structural funds, and occasional employees who had the authority to communicate with appropriate actors at the national or EU level in the context of interest articulation of the city/kraj on these levels (typically in the Offi ce of the Mayor or the Offi ce of the Hejtman 9 ).Other respondents (who functioned as a control group) included employees of corresponding structures at the national level that articulate and aggregate interests of cities and kraje; these structures have means of communicating their interests to appropriate government authorities (typically the executive branch, and especially the Ministry for Regional Development), and in some cases to structures at the supranational level (EU institutions).In the case of cities, this meant the aforementioned Union of Towns and Municipalities of the CR (UTMCR); in the case of the kraje, the Association of Regions of the Czech Republic (ARCR).Cities and kraje can also communicate directly with ministries, so employees of the Czech Ministry for Regional Development were also included among respondents.Due to their professional focus, background and experience, some of the respondents had useful things to say also regarding the relevant structures on the EU level (CEMR, the Committ ee of the Regions, the European Parliament).
Altogether, 16 personal and one telephone interview took place.Information gleaned from these semi-structured interviews was supplemented by follow-up emails when necessary.Interviews (data collection) took place in the period between May 2014 and March 2015, thus in a period in which the circumstances that the individual respondents experienced regarding preparation and implementation of ITI were still fresh in their memory.By respondent request, all interviews have been anonymized.
The Czech case is not only interesting due to the aforementioned implementation of this new tool for territorial development in a de facto unitary state (with a tradition of local autonomy that was interrupted during the 20 th century).Above all, the dis-7 For the characteristics concerning both the appropriateness and also the limits of qualitative interviews as data collection method see e.g., Burnham et al. (2008) or Dexter (2006).8 Some of the interviews were made as a part of broader research, which was focused on the activities of Czech cities and regions concerning the articulation and promotion of their interests in the process of the development and implementation of EU legislation.As the aims of the research were to provide feedback and recommendation for further activities in these fi elds, it was focused mainly on the persons who were directly responsible for the everyday execution of these tasks and activities, i.e., the responsible local and regional offi cials and the representatives of particular structures for articulation and aggregation of the local and regional interests -the Association of Towns and Municipalities of the Czech Republic and the Association of Regions of the CR. 9 The Hejtman is the elected leader of the kraj.
continuity at the regional level makes this case important.Despite this, both types of actors -kraje as well as cities -have developed more than a few initiatives to capitalize on the implementation of ITI on 'their' level, either regionally or at the level of the agglomeration.Beside the parameters of ITI specifi ed (somewhat vaguely) at the EU level, specifi c internal characteristics (such as territorial administrative structure, the positions of individual actors in the negotiating process, and the motivations of these actors) also aff ect how ITI is implemented in the CR.

A discussion on Czech ITI implementation
Since about 2012, an intense discussion has taken place in the Czech Republic regarding the implementation of the 'Integrated Territorial Investment' program, one tool unveiled in the context of 'territorial and urban dimensions of cohesion policy 2014+'.The European Commission explained that 'economic and social cohesion cannot be achieved at the European level without a stronger focus on the territorial impact of EU policies' (European Commission, 2014, p. 2).
One problem since the very beginning has been the vague defi nition of the territorial units in which ITI is to be implemented.The basic conditions the Commission refers to are limited to stating that implementation should occur in 'a designated territory and an integrated territorial development strategy' (European Commission, 2014, p. 2).This was followed by another explanation, which also was not very precise: 'Any geographical area with particular territorial features can be the subject of an ITI, ranging from specifi c urban neighborhoods with multiple deprivations to the urban, metropolitan, urban-rural, sub-regional, or inter-regional levels.An ITI can also deliver integrated actions in detached geographical units with similar characteristics within a region (e.g., a network of small or medium-sized cities).It is not compulsory for an ITI to cover the whole territory of an administrative unit.' (European Commission, 2014, p. 2).
From the perspective of the European Commission, this is a relatively understandable step, as the territorial administration of Member States in the EU is varied enough that any more precise defi nition under which ITI might be implemented would be problematic.At fi rst glance, it was not obvious whether future ITI implementation would be dealt with more at the regional or at the local level.Though clearly the majority of the defi nitions discuss various forms of subregional forms of territory, it is crucial to note that it is not 'compulsory' for an ITI to 'cover the whole territory of an administrative unit', though of course it implies that it theoretically could.From the perspective of the typology developed by Arthur Benz (2009, comp.Marks and Hooghe, 2005), ITI itself is evidence of the so-called 'functional federalism' of the EU, which is characterized by the formation of 'special-purpose jurisdictions'; in the case of ITI, subnational entities have a single goal: to implement ITI.The formation of specialized jurisdictions having only a single specifi c goal is a logical phenomenon, as many territories that have been formed by history and geography are ill-suited to fulfi ll the conditions needed for more modern undertakings, and it is thus necessary to create ad hoc territorial units.The eff ect of the Commission's relatively open design of ITI thus was a situation in which a wide variety of actors started to realize that they could implement ITI, and went about att empts to fully maximize these options.
In the Czech Republic, urban agglomerations and especially the kraje made attempts from the outset.In the period 2007-2013 the kraje had 'Regional Operational Programs (ROPs)' at their disposal; the media (and not only them), however, associated these with corruption scandals.Eventually, key players in some kraje were arrested, and EU regional policy became a shorthand for corruption in the eyes of the public (see Havlík, 2015).The government responded to these developments by deciding to scrap individual ROPs for 2014-2020 and form a single 'Joint Regional Operational Program (JROP)' which would centralize the management of regional policy and would (in the government's view) guard against possible continued corruption on the regional level.Thus at the start of the 2014-2020 round, kraje faced a situation in which a lucrative, fi nancially secure, subsidized territorial instrument was clearly welcomed.Moreover, they were able to point out the fact that many other EU countries were considering ITI implementation exactly at the regional level (Romania, Spain, etc.), or in both regions and urban areas (Slovakia, Italy, Germany, etc.).For a long time, this argument seemed decisive (or at least the kraje were convinced).
Major cities, however, also made an argument that they, rather than peripheral areas (with reference to reducing regional disparities) should be supported as 'poles of growth' (see e.g., Zimmermann and Heinelt, 2012; Hradec Kralove, 2014); major cities along with their agglomerations had the potential to be engines of economic growth and job creation.It was this point that substantially divided the arguments of the kraje and the cities, and created short-term, mutually antagonistic postures.

Divided strategies of Czech kraje and cities in the process of interest articulation
Options and strategies of cities and kraje in the CR often diff er in the interest-articulation process.By defi nition, one reason for this is that kraje are geographically and demographically larger and thus have a logically greater chance of being 'heard'.In the context of EU regional policy implementation, they have enjoyed a more advantageous position than cities in the last decade.They have been administrators of ROPs (some of them de jure, the others de facto, see above) and more frequently represented in key advisory bodies, including ROP monitoring committ ees, ministerial working groups, etc.In these same fora, cities have usually had only one or two representatives, the same position as NGOs.Kraje thus have been in a bett er position vis-à-vis cities from the point of view of access to information and options to present their interests.
Cities and kraje share some basic mechanisms when discussing interest articulation.Both types of actors have umbrella organizations whose goal is to aggregate the interests of their members, select a common position, and accordingly articulate it to other actors.In the case of EU regional policy, these 'other actors' are typically the government or respectively the relevant ministry; at the EU level, typically the Council of European Municipalities and Regions (CEMR), Eurocities (for major cities), the Committ ee of the Regions, the European Commission, etc.Both national umbrella organizations -the UTMCR and the ARCR -are fully professionalized and completely aware of the options they off er; their diff erences can be seen in (A) how they utilize their options, and (B) which communication channels they choose to articulate their interests.
In 2000, the government took the step of creating the kraje as substate units.However, the main motivation was not to counter any trend of centralization, but instead to prepare the country for EU membership; a functioning regional administration was one condition of membership (comp.Dočkal and Kozlová, 2006).This is important also from the point of view of the overall development of political and administrative culture in the Czech Republic, when we think about the position of substate (in this case regional) actors in the whole system of communication, policy making, and governance in the CR as a unitary state.Kraje thus from the beginning of their existence should be understood as new actors (in spite of the above-mentioned 'history' of communist kraje), who had to create their communication channels from scratch.
They never completely managed to emancipate themselves and present their interests on EU aff airs except mainly through the central government, which they considered (and still in part consider) as their natural and closest partner in spite of their status as autonomous substate actors.The interviews indicate close communication on the part of the kraje with ministries and conversely their hesitation to establish or reestablish their representatives in Brussels. 10These kraje focus on articulating their interests through national-level access points, especially the ministerial level, which they also see as a natural partner for interest promotion at the EU level. 11his may be illustrated also by the fact that several Czech kraje terminated their direct representation in Brussels in recent years.The reasons were not only economic, but also personnel (diffi culties to fi nd proper staff with enough experience in working on various levels of governance, including not only regional or national, but also the EU level).These facts in combination with frequent fl uctuation of this properly trained and experienced staff represent certain negative elements of the administrative culture, which are not without impacts on the practice and eff ectiveness of the process of interest articulation on the supranational levels ('[i]t's all about who is sitting at the permanent representation in Brussels.If they're only there to be there, then it's meaningless.'Interview, Kraj Karlovarský, 10 September 2014).It is necessary to point out that the experience with permanent regional representation in Brussels varies in diff erent Czech regions.Nevertheless, the kraje which do have an offi ce in Brussels see it primarily as a fi gurehead position, as typically have one employee (and possibly a single intern), which is not suffi cient to systematically promote the interests of the kraj in the EU; in addition, it makes this activity highly dependent on the personnel factor. 12Even 15 years after its founding, the Association of Regions of the CR has not yet established a common representation of the kraje in Brussels, though respondents of the Association indicated this is regularly considered (according to ARCR, a common representation of the kraje in Brussels would help unite capacities and end the current fragmented situation).
Cities had been represented in Brussels through the UTMCR since 2007, though this offi ce was later closed.In this case, however, it was understandable.Oldřich Vlasák, the vice-chairman of the UTMCR, has been the executive president of the CEMR since 2002, and in 2012-2014 was a vice-president of the European Parliament.Interests of Czech cities therefore have been represented at a high level and any other advocacy was unnecessary in the UTMCR's view. 13 comparison of cities and kraje shows that on the national level, kraje use similar communication channels to cities; however, the kraje are bett er represented on key committ ees (see above) and working groups, allowing them to bett er communicate with the executive at the national level.On the other hand, on the EU level, cities benefi ted from personal contact with the executive president of CEMR and a vice-president of the European Parliament.In spite of the fact that at fi rst sight the position of cities seemed less comfortable on the national level, this could be transformed into an advantage if we take into account that they could further benefi t from a longer 12 On the other hand, kraje that see direct representation in Brussels as useful emphasized its importance for gaining information and contacts ('[the kraj's representation in Brussels] keeps its eyes open, gets involved, watches for chances for funding, deals with various activities in Brussels that occur, which could help or harm the kraje and especially South Moravia, gives presentations at those [...]') (Interview, South Moravia, 25 July 2014; comp.Sălăgeanu, 2014).13 From the long-term perspective, this tactic also could have been seen as rather shortsighted, as it again connected to a great extent the eff ective promotion of the interests and positions of the cities with a particular person possessing both the experience and contacts on infl uential national and EU fora.This again only illustrates the limits of strategic behavior in an environment of limited fi nancial and personal resources and in an atmosphere of a political culture that is not used to putt ing enough emphasis and importance on the communication between the substate and supranational level and the process of interests promotion via these channels.
experience with interest advocacy.In other words, they were in the habit of being actors that had to fi rmly articulate their interests and in this process utilize a wider spectrum of contact points.Naturally, this is again not true for Czech cities en bloc.An important role is played by the municipalities themselves that take an active interest in EU legislation even at the preparation stage on the level of the relevant EU institutions, and not just when it is implemented into the Czech legal system.This means not only having the ability to prepare eff ective communication strategies with the relevant points on the national level (especially the UTMCR, or a ministry) and use of available channels at the supranational level, but also the availability of appropriate structures and (again) especially personnel in specifi c offi ces.
In the case of the new Member States, it is true that at the (not only) substate level, the development of a team, which ideally would have extensive experience from multiple levels of representation of local and regional interests (city/kraj, UTMCR/ ARCR, possibly ministerial positions, or even direct experience with EU institutions, particularly the Committ ee of the Regions), is a long-term project. 14The appropriate structures to deal with substate EU issues (especially in connection with the structural funds) were created in the context of EU accession, and have been understaff ed, especially at the beginning.Eff ective work with European Union issues, which in addition to simply obtaining information also includes eff ective interest articulation of substate actors by means of various available channels on the national and supranational level, requires continuity of personnel; this is critical for the building of formal as well as informal contacts, which play also a vital role (especially) at the supranational level ('I think that in this area there are reserves.It won't do to leave out the horse-trading which took place after the last elections and still continues at the ministries.That's also not completely OK.' Interview, City Hall of Olomouc, 4 July 2014).It is also important to be aware of the importance of the EU issues from the side of the relevant players ('Unfortunately, other countries understand this reality bett er.If you want to promote some interest on the Union level, you also have to be a benefi t for them.And that's not just true for the Council.' Interview, City Hall of Plzeň, 24 July 2014).
14 Interviews with representatives from cities as well as kraje stressed the importance of building internal structures and communications regarding the relevant institutions at the substate level ('[a] mechanism can be built well, but the problem might be communication in the context of an individual kraj and the Regional Authority.The ARCR can't control how information internally fl ows through the kraj.Each Hejtman [the elected leader of a kraj] gets the information, but the question is how it then gets transmitt ed to lower levels.'Interview, Association of Regions of the CR, 9 July 2014).The structural changes in the form of establishment of particular units or positions on the municipal and regional level were rather the outcome of top-down Europeanization in the polity dimension.Personnel discontinuity (and in some cases also insuffi ciency) has been another frequent limiting factor for building long-term contacts, which are necessary for eff ective lobbyist activity, especially on the EU level.

ITI as a divisive factor
ITI has been divisive, escalating tensions between cities and kraje since the very beginning of the discussion about how to implement this instrument, around 2012.While kraje att empted to gain an advantage in new ROPs, which had yet not been promised to them for the following programming period, cities hoped to use ITI to strengthen the growth potential of major areas of agglomeration.Respondents generally agreed that from 2012 to January 2014 the kraje counted on them implementing ITI and altogether receiving several billion euros.The position of the kraje is understandable, insofar as they had become used to the idea that EU Regional Policy was something that had always been implemented at the level of regions up to that point (with a few exceptions), and that they could not imagine it being any other way.However, it soon became clear that the government was more inclined to the development of urban agglomerations.In particular, this inclination was confi rmed by the 'Regional Development Strategy 2014-2020', a key strategic document regarding regional policy that emphasized urban development.The Strategy noted that '[i]n relation to urban issues, there is an increased role of cities in the system of implementation (e.g., delegating some tasks and responsibilities of the administration to cities, strengthening of partnerships and strategic planning, integrated approaches).The second level thus is increased emphasis on support of satisfying development needs at the subregional and local levels [...].ITI will be a prominent implementation tool in the sections of development plans for metropolitan areas and the largest urban agglomerations' (Ministry of Regional Development of the Czech Republic, 2013, pp.109-111).From this point on, major cities' hopes grew for ITI; nevertheless, kraje continuously tried to convince the government, and were far from losing hope in January 2014 during crucial negotiations among all actors.
Municipalities focused their strategy on another logical contact point -the European Commission -which is a key actor in the process of negotiating the form of implementation of regional policy.It is the Commission who signs the Partnership Agreement with the government of the Member State at the end of the process.Cities were personally assured that the Commission never considered that kraje would implement ITI, and repeated this stance during questioning by the UTMCR two months later.It was exactly at this point that the kraje undertook no meaningful activity at the EU level, and based on similar answers from key respondents, probably did not even know about the meetings with cities (based on the interviews, i.e., City Hall of Brno, 22 July 2014 or City Hall of Ostrava, 8 July 2014).It was at this time that the kraje likely suff ered most from the lack of permanent representation in Brussels (in contrast to regions of many EU states who send representatives), as well as from their reluctance to think in terms of lobbying and multi-level governance.
This leads to the question of why at such a critical phase of the negotiations, the respective ministry stood up for the major cities and not the kraje.One partial explanation may lie in the political affi liation of the participating actors to which some respondents referred.While the majority of the principals of the kraje were members of the strongest government party, the ČSSD, since 2013 the key post of Minister of Regional Development has been in the hands of the second-strongest party, ANO (also in the government); ANO may have had a problem with allocating such an amount of fi nancing to the kraje, which are controlled by their political rivals.Less 'conspiratorial' is the argument discussed above.Around 2012 or 2013, the government was already worried about the implementation of ROPs, as these were associated with an increase in corruption.Thus the goal was likely to exclude the kraje and the regional level as such from fi nancial decision-making in regional policy.Therefore, from a certain point of view, the ITI can be seen as a 'welcomed' tool for the respective Czech Ministry of Regional Development to prevent the kraje from implementing the EU money.This clearly shows how the multiple opportunities the ITI off ers can lead to possible confl ict between old and new regions, especially in the specifi c context of EU Member States that had negative experiences with implementation of the EU funds as a result of corruption scandals.(We do not argue that this development is typical for the newer, post-communist EU member states, but the lack of experience and political culture do play a role here, and this should not be underestimated).
Since the Commission did not clearly say which actor should implement the money, national state administration shifted into that position and thus got a powerful tool.In this environment, the cities managed to fi nd their way around and took advantage of the situation by turning to the supranational level channel, in spite of their initially less advantageous position in the negotiations and communication on the national level.Of course, the factor of internal political changes (as mentioned above) must also be taken into account if we want to fully understand the factors which may have infl uenced the governmental decision.
From the point of view of the current understanding of actors' interest articulation in the context of EU regional policy in the CR, we fi nd a relatively new image of mutual relations.Cities and kraje were faced with a situation in which they had the same goal, but only one of them could succeed.Obtaining the chance to implement ITI was strategic, particularly from the perspective of the substantial amount of money which could be budgeted to implement this tool in 2014-2020.Throughout the negotiating process various factors played a role as we described above.However, the strategy of cities focused on promoting their interests directly at the EU and the Commission respectively; this was a step that the kraje did not take at a crucial point.It could serve as an example of the importance of direct lobbying at the EU level, which not only the substate actors in the CR often have tended to underestimate.

Conclusions
Regionalization as a trend of the last two decades, particularly with the support of the European Commission, brings challenges to the concept of multi-level governance regarding precision.As more and more att ention is paid to substate actors, it is increasingly important to diff erentiate among them.This is logical particularly when considering that the European Commission has shifted from its original emphasis on regional implementation of structural funds since 2000, and has increasingly looked at the metropolitan level, which is a degree 'lower' in MLG.As a result, it is increasingly appropriate to look at MLG as a four-level concept, dividing the local and regional levels from each other.Hand in hand with this process we now see a transformation in the character of governance structures, whereby the original 'substate actors' disintegrate into smaller basic units (cities, agglomerations) that have played ever larger roles in the context of European governance.This generally confi rms the validity of Marks' defi nition from the beginning of the 1990s, which explicitly stressed the role of the local sphere in MLG.
We have demonstrated the splitt ing of these levels by their specifi c interests and preferences in the example of negotiations surrounding the implementation of the ITI regional policy tool in the Czech Republic.It turns out that in the discussion of this financial tool the Czech kraje and cities diverged signifi cantly in their preferences.One key point in the entire process was the ability of cities to recognize their position in the MLG framework and their decision to bargain for a more advantageous position directly at the European Commission.Kraje, on the other hand, chose a conservative, more 'traditional' approach and their strategies, focused almost completely on the centralized authorities in Prague.
In the overall process, kraje made several fundamental mistakes.They: A) underestimated the logic of negotiating on multiple levels and over-relied on working with the government.Further, they probably B) underestimated the political nature of decision-making; at key points, the Ministry of Regional Development sided with cities or agglomerations led by major cities respectively.While kraje have mostly been headed by Social Democratic politicians, the relevant ministry was headed by their political rivals.While it is not possible to remodel history, it can be assumed that in the case of early establishment and maintenance of contact with the European Commission, the kraje could have prevented adverse developments.However, individual actors were not prepared; especially for the kraje, we see that at the beginning of their existence they did not learn the logic of multi-level governance. 15 conclude that the volatile situation between kraje and cities probably would not have occurred in the Czech Republic if the European Commission had not decided to support the so-called territorial dimension of regional policy in recent years.By doing so, whether consciously or unconsciously, the European Commission made special incentive structures (ITI in this case) in order to raise competitiveness and economic growth.The probably unintended consequence of this was confl ict among regional and subregional actors.This can be demonstrated with the case of the CR, where the confl ict took place in the context of problems with the EU structural funds in the previous programming period and where the factor of the development of the internal political situation also played an important role.The outcome of the confl ict then showed the importance of using multiple communication channels in the process of interest promotion of substate actors and may serve as one piece of evidence that multi-level communication should not be underestimated in these cases.
The major argument of this article may be used by further studies describing similar processes in other EU countries as well.The nature of confl ict between new and old regions in diff erent EU member states may depend on their tradition of cohesion policy, as well as on the longterm relations of their substate actors.The confl ict potential, which was created by the new period 2014-2020, may gain various forms both in the older and the newer EU member states.Further research of this topic could bring other case studies, especially of the countries which share similar characteristics of the newer EU member states with the CR, representing the region of Central and Eastern Europe.These characteristics often include the lack of experience with 'multi-level' negotiations of the substate actors, frequently combined also with problems of political and administrative culture.Further research of the described processes both in the older and the newer EU member states may reveal the relative importance of these variables (or factors), which infl uence both the course and outcomes of the confl ict (as we saw in the case of the CR).