Governance and transportation policy networks in the cross-border metropolitan region of Luxembourg: A social network analysis

Introduction

Every workday, approximately 21,000 workers cross the country’s borders to the agglomeration of Luxembourg City by public transportation, which is slightly more than one quarter of the overall number of commuters. The cross-border metropolitan region of Luxembourg is heavily influenced by the economic gravitational centre of Luxembourg City, which, compared to its regional surroundings and despite its modest size, concentrates high metropolitan functions as well as a strong service and, in particular, a financial sector, with the latter accounting for 63,000 direct and indirect employees in 2010 (Luxembourg for Finance, 2012). The agglomeration of Luxembourg City attracts the necessary qualified labour from abroad, most of them living in the bordering regions of Belgium, France and Germany. The region expands per definitionem beyond administrative borders. Hence, physical mobility by road and rail is the key for intensifying the cross-border metropolitan region’s functional integration. Luxembourg’s national transportation infrastructure is part of a larger inter-national road and rail network in the heart of Europe, placing efficient cross-border governance solutions within the public transportation policy field at the centre of attention.

This is in line with the intended regional harmonization process and the political establishment of cross-border regions across the European Union (EU). The setup of the European Spatial Development Perspective (European Commission, 1999) prioritized specific spatial issues. First, a strong emphasis was placed on the infrastructural enlargements of transportation networks to enable and connect parts of different EU regions with each other, among others, financially supported by EU programmes (Hajer, 2000). Second, there was a push in favour of polycentric urban regions as the most efficient scale to implement territorial cohesion policies to prevent centre-periphery predicaments (Newman, 2000; Scott, 2002). Third, border regions received particular attention and funding (INTERREG programmes), which focus on the transformation of the nation states’ territorial peripheries into regions with common economic interests and cooperation (Anderson et al., 2003; Brenner, 1999; Newman, 2003; O’Dowd, 2003; Perkmann, 2003). This redirection of EU spatial policies is further based on an encouragement of mobility (Jensen and Richardson, 2004) to overcome physical and mental barriers, to enable free flows of people, capital, goods and services, and to raise a cross-border region’s international competitiveness. European policies are designed to encourage the rationale to resolve economic, social, and spatial inequalities based on interactions and exchanges between different places (European Commission, 2010). Consequently, the cross-border regionalism is an integral part of the proclaimed de-bordering process within the EU (Diez, 2006; Nelles and Durand, 2014). While the political will at the European level has been shaping cross-border regions for decades, the criticism on an operational level, however, remains that ‘states are generally unwilling to hand over portions of their sovereignty and political authority to the structured forms of co-operation, sometimes prohibiting and frustrating direct and efficient dialogue between partners in the border regions’ (Van Houtum, 2000: 66).

Given the premise of European integration, while also drawing on Van Houtum’s (2000) criticism, we aim to analyse the situated governance structure of cross-border public transport provision on the example of the respective cross-border policy network in the Luxembourg metropolitan region. In our analysis, we go beyond the common politically-predefined territory of cross-border cooperation. With a social network analysis (SNA) approach, we combine a relational measure of the organizations’ centrality with a measure based on collective perceptions of the organization’s powers in terms of decision-making. The unit of analysis is therefore not the politically-established cross-border region ‘Luxembourg and the Greater Region’, but rather a functionally integrated cross-border space of cooperation. Thus, this relational space has no predefined territorial boundaries, as its limitations are set by the functions, actions and the operating distances of the identified network members. For this reason, we refer to a broader defined unit of reference, the ‘cross-border metropolitan region of Luxembourg’ (CBMR). It consists of parts of the four nation states of Belgium, France, Germany and Luxembourg, in which Luxembourg City is the gravitational locus for the massive daily cross-border commuting activities. Although the CBMR’s different actors are increasingly engaged in the specific policy field of public cross-border transportation that concerns the CBMR as a whole, there is evidence that decision-making processes are still fragmented and deeply anchored in traditional national decision-making structures and rationales.

We suggest that the governance typology of Hooghe and Marks (2003, 2010) and Marks and Hooghe (2004) offers a useful guiding heuristic. The discussion on their ideal type II governance shows that such function-specific governance arrangements have indeed emerged in cross-border regional contexts. The empirical findings of our case study, however, seem to contradict this conceptual notion as the general-purpose jurisdictions continue to hold responsibility for public transport provisions. This prompts us to ask to what extent the ‘border effect’ is still resilient and whether ‘borders’ and ‘border effects’ truly limit regional cross-border co-operation.

In an attempt to provide empirically informed answers to these questions and to reveal the prevalent complex nature of situated policy network structures and relations of the CBMR’s dissimilar actors and their operating range, we first introduce two main bodies of literatures – border(land) studies and governance studies – our arguments draw on. The second section introduces the utilized method mix of quantitative and qualitative approaches and technically describes the empirically-drawn policy network with the help of a SNA. In the third section, we explain and critically discuss our empirical results, while we aim to contribute to the conceptual debates on governance and border studies in the politically-proclaimed de-bordering regions within the EU in the concluding section.

EU metropolitan regions beyond ‘borders’

During the past decade, a number of EU border regions have undergone a remarkable transformation from limiting barriers to actively designed policy spaces. The complexity and dynamics of such border regions, however, are still strongly rooted in the logics of the Westphalian sovereignty of nation states and societal power relations (Newman, 2003, 2011; Paasi, 2011). Scholars substantiate that (national) territorial policies of ‘ordering, bordering, and othering’ (Van Houtum and Van Naerssen, 2002: 126) portray borders as the ‘skin of the state’ (Megoran et al., 2005), often rooted ‘in powerful cultural, economic and political inequities’ (Nicol and Minghi, 2005: 681), which cause the separating effects of borders to persevere.Even within the EU – in economic terms, the world’s largest single market and with regard to its encompassing Euro currency a borderless economic entity – national particularism and territorial anxieties still determine the daily agenda (Scott and Van Houtum, 2009) and reinforce hindering ‘border effects’. Yet, some cross-border regions have managed to create ‘hybridity’ (Newman, 2011: 33), tying in with the idea of the ‘border’ as a reflection of a positive meaning to seize opportunities unavailable in their respective nation states (O’Dowd, 2003). With reference to our case study, we argue that ‘border’ is of particular importance for the network actors’ relations and that empirical evidence derived from a SNA would show dissimilar results in non-border regions. National borders are institutions and, thus, concretized policies (Paasi, 1996).

Marks and Hooghe (2004) have argued that a new – admittedly ideal – type of territorial authority (type II governance) has emerged, in which the multiplicity of specific functions is conceptualized in policy networks and ‘multi-level governance’ (Marks, 1993). It complements the traditional, durable centralized authority (or type I governance) of non-intersecting and nested territory. Multilevel governance policies, with their task-specifity, intersecting membership, flexibly-designed competencies, and intervention mechanisms, can be regarded as respective organizational response strategies to the process of ‘Europeanization [as] the quintessential case of state glocalization’ (Nelles and Durand, 2012: 3). Regional integration theories brought the idea of multilevel governance to the fore, defined by an intensification and flexibilization of relations as well as by the increased interactions of actors on different spatial scales. Beyond the supposition that governance in general has become multi-jurisdictional, or multi-layered, involving actors from different administrational scales in the recent political decentralization processes of re-scaling, the custom-designed form of type II governance has not yet been empirically verified within explicit cross-border governance contexts and their policy networks. Although there has as yet been no consensus on how it should be organized (Hooghe and Marks, 2003, 2010), the formulated two ideal, although contrasting, governance types I and II offer great potential in the analysis of cross-border governance contexts, in which ‘a web of transnational jurisdictions’ (Hooghe and Marks, 2010: 25) has been shaped within the course of time. Besides such jurisdictional webs, we have identified groups of very different actors with varying tasks, responsibilities and coverages, who are more or less tightly embedded into their jurisdictions of origin. Deriving from Hooghe and Marks’ typology and applying SNA to the cross-border context at hand, we assess two research questions.

The first research question explores the extent to which, in a type I governance regime, we would expect the governmental actors of their respective national jurisdictions to be the most important decision-makers in the policy network. The background for our first research question is that, as a general rule, public transportation infrastructure is typically costly but – despite EU integration policies – subject to strategic national interests and decisions. However, national public finances deteriorate and EU funding is limited. Because cross-border regions usually do not command their own budgets, the majority of such infrastructural projects can only be realized by a decision backed and financed on a national level (Modellvorhaben der Raumordnung (MORO), 2009). In the context of the CBMR, the careful cost-benefit weighting by the involved states puts practical questions, such as which state pays for and receives what, how much and why in return, on the daily agenda. It can heavily delay the necessary cross-border infrastructure projects.

We therefore suggest that the new phenomenon, which we discuss here, is a strong but twofold competition setting: first and foremost, between national jurisdictions in the cross-border context. The competing jurisdictions in the cross-border setting at hand mirror the prevailing cross-border contradiction of the proclaimed borderless regions (political rescaling) as well as the respective national interest in decision-making. Thus, we still depict the current snapshot of the CBMR as a space being ‘disrupted by the sovereignty of the governments ruling on each side of the frontier’ (Council of Europe, 1995) but slowly transforming into an ‘active’ space and a key area for cross-border policy development (Van Houtum, 2000: 64). This leads us to the second setting of competition we can identify in the case at hand: the one on the organizational transition from the governance types I and II, or, in other words, the assumed breaking of the governance type I dominance. Public transportation is one of the predestined policy fields, which is claimed to be prone to type II governance forms, especially in cross-border formations (Hooghe and Marks, 2010). Following this lead, Nelles and Durand (2012) discern major differences in the governance structures within their comparative assessment of the two European cross-border regions, Lille-Kortrijk-Tournai Eurometropolis and the CBMR. They seek to explain them with the varying historical origins and the founding actors’ perception of the governance working logics. Whereas, for example, the Lille region had been following a bottom-up process, in which ‘the initiative and the institutionalization of cooperation originated from local authorities’ (Nelles and Durand, 2012: 10), the CBMR of Luxembourg’s governance structure was established top down, and the involved states are actuating cooperation rather than a looser facilitation and supervision purpose as is the case in the Lille region.

Connectivity and physical accessibility are crucial for the CBMR’s cities’ attractiveness and competitiveness but their capacity to take action is handicapped by their lack of legal competencies in this policy field at a cross-border scale. The cities have a clear interest in attracting investment activities and are assumed to be crucial information brokers in the process of negotiation and decision-making. However, connectivity is not solely about hard infrastructure; it is also based on important services to ensure a smooth operation of cross-border public transportation. This comprises issues such as tariff harmonization, the merging and synchronization of the bus and train time tables on either side of the border, or the online provision and update of the respective travel information. All of these issues are often overlapping parts in the cross-border transportation policy field involving different – public and private, state and non-state – actors from the four countries. Commonly, they are organized in functionally-oriented and temporarily defined project structures. In conclusion, the above mentioned specifics meet the suggested idealized type II governance.

Our second research question thus asks, to what extent, in a type II governance regime, do non-public actors impact on the decision-making process of central and local states? In other words, to what extent do they perform their ability to actively shape the communication flows and to be embedded in the actors’ network? Yet, the jurisdictions strongly differ from one country to another. The CBMR on the German side extends to the western parts of Rhineland-Palatinate (RP) and Saarland, which are sovereign constituent states (Bundesländer) of the Federal Republic of Germany. In the policy domain of regional public railway transportation, both the respective transportation ministries of the Länder RP and the Saarland hold responsibility (Regionalisierungsgesetz from 1996). The German supraregional bus line system is even more complicated (Beck, 2012). In France, the regional railway competencies (excluding the international high-speed trains) have been progressively transferred from the nation state to the regions in a decentralization process (Loi relative à la solidarité et au renouvellement urbains in 2000). Hence, the Région Lorraine now organizes the regional public railway transportation. In Belgium, the regionalization process transferred the competencies in regard to the organization of the local transportation, i.e., bus lines, from the nation state to the regions in 1988, while the federal government remained responsible for the national railway. From the perspective of public transportation, the competing interests of the Flemish and Wallonia regions, however, tend to prevent continuity of the strategic policy towards Luxembourg. In Luxembourg, railway competencies are a national matter, whereas the bus lines benefit from cross-funding between the communes and the nation state. In general, public transportation is heavily subsidized in Luxembourg, as well as in France and Belgium (Table 1); in Germany, however, inter-urban rail services with a radius of more than 50 km are subject to free market competition. The latter fact creates insurmountable difficulties, e.g., in terms of a common cross-border regional public transportation tariff system, which, on the German side, includes, for example, the connection between Trier and Luxembourg City. The ongoing negotiation process over the past 12 years has not produced acceptable results for the daily commuters, which is one reason for the low number of commuters from the German cross-border regions using public transportation. The example of the single-track bottleneck bridge in Igel (RP) – the long-term subject to intense but ineffective negotiations between Luxembourg and Germany – has been preventing the achievement of the aspired travel time decrease and the enhancement of the share of public rail transportation since the 1990s.

OPEN IN VIEWER

Table 1.

Complex responsibilities in the regions involved in the cross-border transportation policy network in the CBMR of Luxembourg.

	Cross-border train connections	Cross-border bus connections
Belgium	The whole railway policy is run by the Nation State.	Bus lines are run by the Regions and co-funded by Luxembourg.
France	The Central State runs international transportation; Regions are responsible for regional transportation.	Bus lines are run by local transport agencies publicly co-funded by Luxembourg.
Germany	The national and international railway policy is run by the Nation State; the Länder are responsible for regional lines.	Private companies operate on behalf of local public transport authorities (tendering process).
Luxembourg	National railway policy is the jurisdiction of the Central State.	Bus lines are run by private companies co-funded by the State.

To summarize, the observed mismatch in public transportation between the four countries seem to have contributed to an extensive negotiation process in a complex mélange that comprises competing inter-governmental responsibilities as well as national strategic interests regarding large (and expensive) infrastructure projects. Hence, the CBMR of Luxembourg offers a rich empirical example of the stakes and difficulties that underpin the governance building on cross-border metropolitan region’s public transportation themes. In 2004, the state of Luxembourg implemented a spatial planning strategy (Integratives Verkehrs- und Landesentwicklungskonzept) to increase the modal split of public transportation to 25% of the total number of daily cross-border trips. Although the number has increased during the past three years, the score of the daily cross-border commuting flows only amounted to 14% in 2010 (Schmitz and Gerber, 2011). Given the country’s small size, its large dependency on the labour markets in the adjacent regions and the need to channel the huge daily commuter flows as well as its deep embeddedness in the European transportation infrastructural network, Luxembourg’s national public transportation policy cannot be assessed and pursued in isolation from the adjacent countries. This, in turn, exposes cross-border public transportation in the CBMR to an efficient governance model. The example of the cross-border mobility scheme (SMOT) between Luxembourg and France aptly illustrates that – in addition to the implementation within a national context – a national mobility strategy also has to actively consider the conditions in the neighbouring regions. The numbers in Figure 1 depict the importance of public transportation across the adjacent country regions.

OPEN IN VIEWER

Figure 1.

Daily commuter flows towards the agglomeration of Luxembourg City within the CBMR of Luxembourg (2008).

Methodology

The analytical framework is based on a comparative analysis between organizations located in parts of the four nation-states Belgium, France, Germany and Luxembourg, and integrated in the same cross-border policy network. The SNA approach builds on the supposition that relationships among interacting actors are crucial, that actors and their actions are interdependent, and that because of their relational ties they are able to channel flows of (in)tangible resources (Wasserman and Faust, 1994: 4).

A questionnaire of both open- and closed-ended questions guided through 41 face-to-face interviews with selected experts; 12 rejected the interview request (Table 2). For the subsequent SNA, only actors nominated by others at least three times were considered. Besides their attributed characteristics as experts, the interviewees were also nominated members of the respective policy network. The 41 experts interviewed belong to 34 organizations. Large organizations generally have various subdivisions, where sub-decisions regarding a larger project or policy field are made. In cases where the interviewees belonged to the same umbrella organization, i.e., a ministry with subdivisions covering different aspects of cross-border transportation issues, the individual experts’ answers were merged and the answers given by the more senior level person were given priority. However, there were no such cases with truly contradictory answers from staff of different departments.

OPEN IN VIEWER

Table 2.

Overall response rates on the application of the snowballing technique.

	Planned	Not relevant/Not existing 1	Refused	Interviewed	Response rate
Wave 1	44	2	8	34	34/42 (81%)
Wave 2	12	1	4	7	7/11 (63.6%)
Total	56	3	12	41	41/53 (77.4%)

1

The individuals and organizations targeted but not relevant or not existing anymore were not taken into account for the calculation of the response rate.

In the following sections, the paper calls the considered organizations ‘actors’, not any longer referring to individuals and their perspectives. The closed-ended questions included a number of tables and Likert-scale questions and formed the basis of the subsequent SNA. In contrast, the open-ended questions served the purpose of better understanding the different experts’ assessments on achievements, development, satisfaction, further challenges and other sensitivities within the cross-border policy field of public transportation, while also helping to broaden our knowledge about the involved organizations and their motives on the observed governance trade-offs. Thus, the SNA applied here goes beyond the classical modes of observation of governance structures in cross-border policy networks. Concurrently, it also backs up our quantitative results with a more explanatory qualitative analysis. The data covers the network and policy situations between 2008–2009 and 2010–2011 due to cross-border co-operation in public transportation that was predominantly project-based.

One of the major difficulties in SNA is to define appropriate ‘boundaries’ within the investigated network (Freeman, 2004; Scott, 2000). Similar to approaches in other empirical social contexts (Fainstein, 2001; Lai, 2006), we selected the network actors by applying a ‘snowballing’ technique, or ‘reputational method’, in which interview partners nominated other network members whom they perceived as important from their point of view (Table 2). Due to the high turnover of individual network members, but also owing to a lack of updated network member lists, a two waves’ snowballing technique (Christopoulos, 2009) was performed, which practicably allowed for the identification and nomination of the essential organizations and actors in all four of the involved CBMR’s countries.

Our empirical findings are based on the exploitation of two kinds of measures. The first one reflects the classical SNA measure of the normalized degree of centrality. It expresses the number of an organization’s links with other organizations within the network, divided by the total number of potential links (Freeman et al., 1991). Accordingly, its value is between 0 (no tie at all) and 1 (the organization is linked with all of the other ones). This measure reflects the multiplicity and diversity of actors an organization exchanges information with, but does not give information about the intensity and relevance of the exchanged information. It allows the identification of the organizations that are involved in the governance and actively participate in the communication structures by sending and receiving information. As a consequence, we utilize this indicator to analyse the second research question, in which we investigate the positioning of the non-state actors in the context of information flows.

The second indicator is the perceived level of importance in the decision-making process among the different network organizations. We call this indicator a ‘decisional power index’. This measure is based on the results of a closed-ended question that all the interviewees were asked: Who are the most important actors in the field of cross-border public transportation? The decisional power value for an organization is obtained by adding the total number of times the individuals who represent this specific organization were mentioned as being among the most important actors in the policy field. The total number for all the individuals who represent an organization indicates the result for each organization, considering that ‘the social capital of people aggregates into the social capital of organizations’ (Burt, 1992: 9). This information is crucial to differentiate between the organizations that share information without having the power to take final binding decisions and the ones that effectively lead the decision-making process.

To reduce the heterogeneity and to generalize our empirical findings, we clustered the different surveyed organizations in typological homogeneous groups (Figure 2). For this reason, we applied the method of the empirically-grounded construction of types and typologies (Kelle and Kluge, 2010), with its four central analytical steps of 1) developing relevant analysing dimensions; 2) grouping the case and analysing empirical regularities; 3) examining meaningful relationships and constructing types (Table 3); and 4) characterizing the constructed types (Kluge, 2000). These indicators allow us to validate the first research question, in which we test for the power of state actors towards the implementation of concrete political decisions.

OPEN IN VIEWER

Table 3.

Typology of involved cross-border governance actors.

Category (number of actors)	Degree of decisional power	Normalized degree of centrality
1) ‘Leaders’ (2)	++	++
2) ‘Secondary decision-makers’ (5)	+	+
3) ‘Information diffusers’ (25)	−	+
4) ‘Marginal actors’ (32)	−	−

OPEN IN VIEWER

Figure 2.

Empirically grounded types of network actors for the CBMR of Luxembourg.

The represented network (Figure 3) does not consider the direction of the relations between the actors mentioned by the interviewed experts. Therefore, we have symmetrized the matrix of ties, assuming that, if an interviewee claimed that s/he had links with another person/organization, this was true, even if this exchange of information was not confirmed by the interviewees in the other organization. In fact, the purpose of the study was to underline the current stage of the process of information sharing, which is given by undirected links, whereas directed links are used to highlight who seeks advice from whom (Hoppe and Reinelt, 2010). This allowed us to avoid neglecting actors when experts forgot to mention established relations to other network actors, although their counterparts did.

OPEN IN VIEWER

Figure 3.

Cross-border public transportation policy network in the CBMR of Luxembourg.

Unravelling the intricacies of policy-making: A resilience of traditional governance patterns?

According to the dimensions of the two analyses, decisional power and centrality, the cross-border policy network suggests the following typology of the involved actors (see Table 3).

Towards an alternative conception of multi-level governance in a cross-border context

These observations seem to validate the first research question but fail to fully support the second, because except the ESCGR, no other non-state actor, e.g. local transport agencies or associations that belong to the identified decision-makers, reaches high enough values of interaction. The prevalent situation in the CBMR for the policy field of public transportation shows that cross-border operative, strategic and decisional tasks as well as the working level of actors on different scales overlap; but empirical evidence strongly suggests that communication structures remain hierarchically anchored and prevailing. More precisely, the differences in the administrative configuration of the local, the sub-regional, the regional, and the national levels in the four member states of the CBMR seem to prevent necessary direct and task-specific communication. Thus, the varying national systems and priorities call for governance solutions among a more heterogeneous group of cross-border actors. In the example at hand, political state-centred traditions, such as in France and Luxembourg, meet federally-organized state systems, such as the ones in Belgium or Germany. In this setting of different, and even competing understandings of the political organization (Perkmann, 2003), decision-making seems to be difficult and pervaded by a number of trade-offs. Empirical evidence suggests that the actors in the analysed policy network indeed meet selected criteria of the type II governance; however, the prevalent governance structure is still that of the traditional type I, state-centred organizational regime. At the same time, actors of the third group – information diffusers – still seem to lack the political resources (or will) to make full use of the potential that the task-specific governance-II form offers, as the following interview informant suggests:

The problem is that the decision-making in regard to transnational transport is basically done on the level of the ministries. This is a disadvantage, because ministries are not directly in contact with the general public. Hence, they should not make decisions alone but in close cooperation with the communities who are actually confronted with these problems every day. (Luxembourgish actor, May 2011)

In fact, this point of view was shared by numerous actors. They themselves call for a more bottom-up organized network of structures and functions. However, on the network’s governance level, it does not yet work. Rather, it shows that political re-scaling processes in the CBMR of Luxembourg are, dissimilar to the afore-mentioned Lille region, still in their infancy. Although the CBMR commands a functionally-specific policy network with influential and diverse regional public and private actors, in which information flows and exchange is the key, a number of interviewees still referred to a strong authority, such as, for example, the nation state, as a necessary precondition to make this ‘flexible’ governance work in the case at hand. The following quote exemplifies this:

On either side, a strong politician is missing, someone who has the political authority to make a difference. The problem is that politicians are always elected for a limited time frame. It’s almost impossible to temporarily have strong actors on all sides who want to move forward and also cooperate well. Great projects in the past have shown that at least two politicians have to act in concert transnationally. If that is not the case, you have to go the long way via all the different authorities. Then numerous actors will be involved, with whom you need to find agreements. And when you’ve reached these agreements, it remains a question of financing. . . . If nobody is keen on promoting the project, someone who can persuade the minister and get his approval if necessary, then you have a problem. (Luxembourgish actor, April 2011)

Aggravating the present situation in the policy network at hand, the negotiated solutions are expensive, time-consuming and ‘incur exponentially rising and eventually prohibitive transaction costs’ (Scharpf, 1997: 70), which poses the question: Are governance modes according to the ideal type II really suitable for this kind of diverse cross-border policy network? It becomes clear that such a network demands leading actors who are able to formulate an overarching vision and design a common strategy. Although there is no general blueprint for the way the governance arrangements in such cross-border regions should be organized, a number of policy network actors did mention a supranational authoritative body could steer the public transportation in the cross-border context and eventually lead ‘… to the establishment of specific governance networks’ (French actor, July 2011).

How can we contribute to the conceptualization of the border situation with our case study? We argue that the ‘border’ is of particular importance for the network actors’ relations; empirical evidence derived from SNA would show dissimilar results in non-border regions. Borders within the CBMR function through a ‘selective openness’ (Paasi, 2011: 12). Social practices of the differently empowered network actors ambivalently maintain or challenge the territorially demarcated spaces. First, the necessities of physical access in CBMRs render public transportation a key for local/regional actors to exploit not only economic advantages but also to mobilize political resources. Hence, the enormous mobility between regions with such dissimilar production factor endowments (strong economic centre, strong adjacent labour resources), as it is the case in the CBMR but also in the border regions of Basel or Geneva (Sohn et al., 2009), weakens the border effect and challenges its resisting social practices and physical materialization. Allowing for a more flexible integration of cross-border public transportation could potentially lead to a win-win-situation for two distinctly endowed border regions: One resolves its lack of qualified labourers, the other – economically much weaker – experiences economic stabilization. Second, the interests of the ‘power elites’ (Newman, 2011), i.e. the political and economic gatekeepers on the different sides of the border, still differ, while their social practices seem to enhance and maintain the dividing border situation. The observable policy practices in the different sub-fields of cross-border public transportation illuminate for instance that, while cooperation among cross-border actors on a working level generally seems to run smoothly and is well rehearsed, this is not always the case on a political level, leaving involved actors at times to frustratingly state: ‘Transnational public transport stops at the border in many respects’ (German actor, June 2011).

The fight for the scarce financial resources between the involved (local and nation) states on the example of large cross-border infrastructure projects brings the border effect to the fore. Further and due to their contradicting competencies in each of the involved national jurisdictions in the CBMR of Luxembourg, ‘the capacity to take collective decisions, and make them stick, is diffused among a wide variety of actors’ (Hooghe and Marks, 2010: 22; Ostrom and Walker, 1997) in the four different member states. Their dissimilar competencies and resources contradict, or at least complicate, cross-border co-operation in the case at hand, and their intersecting memberships are influenced by a number of coexisting, but also competing, centres of decision-making with formal independence from each other. However, considering the overcoming of borders to be a contingent process along a continuum of possibilities, in which the setup of governing border management structures is a series of discourses, negotiations, structuring, and decision-making, the CBMR of Luxembourg may find itself in a transitional stage between the ideal governance types I and II. Thus, we propose a third – hybrid – form of governance (Figure 4), which combines characteristics of both type I and type II governance.

OPEN IN VIEWER

Figure 4.

Key characteristics of type I, type II, and hybrid governance regimes.

An aggregation of both elements, the network’s structure and agency, suggest this hybrid form of evolving bottom-up policy processes. We illustrated that not only cooperation, but also competing economic structures and interests, competing cultural understandings, as well as competing political ideas, shape the complexities of the policy network at hand. Still, the institutionalization of communication, i.e. in the form of the ESCGR, comprises both the private and the public sector and includes regular topic related round tables and working group meetings but also emerging forms of participation of a broad range of actors within the frame of functionally and temporarily-defined projects (e.g. the set-up and harmonization of online information of bus and train time schedules across the CBMR). It elucidates that a strong influence by non-state actors on the decision-making processes in the policy field of cross-border public transportation has been forming, which is yet more complex than a ‘classical’ type I governance but less encompassing than the type II governance. We therefore promote the idea of a hybrid, intermediate mode of governance.

Conclusion

The paper pursued two aims. First, we investigated the additional value of the SNA on cross-border governance networks’ analysis in Europe. Second, we discussed the current stage of such a multi-layered, complex policy network of public transport provision on the specific example of the CBMR of Luxembourg with its many stakeholders in order to better comprehend how cross-border cooperation is structured in this context.

The applied SNA on this particular cross-border governance context allowed us to get a situational analysis of the respective policy network; in other words, we took one picture from a continuing sequence. This ‘frozen image’ has clearly helped us to identify and evaluate the configuration of the network’s (missing) relations and key actors as well as – to a certain extent – the role of the involved network actors. The latter was supported by empirical evidence provided by interviews with the respective network members. This innovative approach provided us with a tool, whose analytical outcomes may help us to better understand the difficulties that constrain the organization of an effective political integration of border areas. Despite the all-embracing European rhetoric and provided financial resources, a number of cross-border regions still face such challenges.

Indeed, we have shown that the presence of state borders and, therefore, competing national solutions of public transportation provision still present smaller or bigger hurdles for efficient decision-making. This even applies to a cross-border context, in which concrete and immediate common problems aim for a unifying motivation of cooperation. The policy domain of public transportation has been presented as being likely to lead to the emergence of a type II governance regime (Hooghe and Marks 2003). However, despite the widely expressed interest for a stronger cross-border policy involvement by numerous local and regional network actors, we have shown that there is still a strong inertia of central state power when it comes to binding project decisions, illustrating that the ‘sovereign state with clear national interests is still alive and kicking’ (De Vries, 2008: 57). In contrast, the precursory negotiation, discussion and mediation processes are based on a much broader foundation of involved network actors and their respective multilateral information flows than it was the case two decades ago.

The policy field of public transportation, especially in terms of transportation infrastructure, is most often of high strategic importance to the nation states. This makes finding a consensus across borders, even within an inclusive economic space, exceptionally complex. In similar settings to the case at hand, such as the border regions around the strong economic centres of Geneva or Basel (ESPON, 2010), political border effects play a crucial as much as a hindering role. The empirical results of the case study at hand therefore suggest ambivalent conclusions. On the one hand and although we would wish to be careful with the generalization of our results, one finds evidence in the mentioned cross-border regions that the institutional ‘mismatches’ between the included state actors with sovereign powers constitute an obstacle towards the proclaimed emergence of bottom-up governance types (for the case of Basel, see Walther and Reitel, in press). It appears that this specific cross-border context is not particularly beneficial to the assertion of new ways of governing and rather seems to reinforce traditional and nation state-centred decision-makings as a result of national jurisdictions competing with each other within the CBMR.

On the other hand, the same cross-border context more heavily promotes the involvement of many additional actors in the phases of negotiation and decision preparation than would be the case in the tradition of nation state-centred decision-making. Information exchange among a variety of public and private actors is well established, and the decision-making cannot be imagined any longer without extensive, regular, and influential up-front communication processes. They have been forming the basis for advocacy and lobbying, for shared interests, for coalition and consensus building, as well as for the development of a shared vision for the region as such. Meanwhile, this is also influencing the decision-makers and the decision-making processes themselves; all of them are contextual elements, which, at best, lead to a stronger formation of a type II governance in the (near) future.

The contradictory notions of the politically intended cross-border policy integration vs. the neo-liberally inspired and politically-encouraged build-up of competitive capacities of each region place special emphasis on the border as a prevailing institution and, thus, on concretized policies with enrooted social practices. The politically sensitive policy field of public transportation may require more specific instruments and incentives to overcome the vested rights on a national level. But even within these complexities and contradicting circumstances, the potential of the ambivalent border perception of the different stakeholders becomes clear: formerly marginalized spaces experience new potential for economic growth and socio-spatial stabilization. Thus, we agree with Paasi (2011: 38) that ‘borderlands can become transformed into zones of transition in regions of stability and cross-border cooperation’. However, this contingent transformation process is actively shaped by the power elites on either side of the border and, thus, by the outlined cross-border policy network governance in its hybrid state. It would be fruitful to test the assumed existence of a hybrid form of governance by comparing our results with other case studies of policy networks in other policy fields than public transportation but also in different geographical contexts.