GERMAN STAKEHOLDER PERSPECTIVES ON THE PROVISION OF CROSS-BORDER PUBLIC SERVICES

Differences in the legal and administrative frameworks of the member states of the European Union (EU) have been identified as presenting major challenges to achieving territorial cohesion. The policy debate culminated in 2018 with a Commission proposal for an EU Regulation on a ‘European Cross-Border Mechanism’ (ECBM). While the proposed legal instrument is still under negotiation, this paper analyses the experiences of stakeholders in Germany’s border regions on navigating obstacles in the provision of cross-border public services. The paper concludes that while more legal certainty for crossborder cooperation would generally be welcome, stakeholders in border regions do consider local and high-level political support; financial incentives to address complex border challenges; and dedicated local and regional ‘policy entrepreneurs’ to be at least as important for realising complex joint projects.


Introduction
Despite many years of institutional and financial support for cross-border cooperation in the European Union (EU), border regions continue to be faced with numerous obstacles to the provision of joint public services. Border regions are commonly defined as those parts of a national territory of two or more neighbouring nation-states that are separated by one or more national borders, with sub-national authorities the key actors for cross-border cooperation (Noferini, Berzi, Camonita & Durà, 2020). According to recent analyses by the European Commission, there are 40 internal borders in the EU, which are home to almost 30% of the EU population and produce 30% of its Gross Domestic Product (GDP) (EC, 2017a). Yet, access to key services of general interest (SGIs), such as health care or universities, is generally lower in border regions as a result of the cost of navigating between different administrative and legal systems (EC, 2017a). Such crossborder public services (CPS) are delivered through existing domestic or new joint infrastructures and therefore differ from project-level 'ad hoc' cross-border cooperation (ESPON, 2020). They are provided with the intention of a long-term basis for the benefit of the population of the border region and often require complex formal agreements between neighbouring states and regions to bridge differences in national regulations and technical standards.
In recent years, considerable efforts were devoted to identifying the main legal and administrative obstacles to realising such more institutionalised forms of cross-border initiatives, and on formulating responses to the challenges of implementing CPS. This work has in 2018 culminated in a European Commission proposal for an EU Regulation on a 'European Cross-Border Mechanism' (ECBM) (EC, 2018), presented as part of the regulatory package for EU Cohesion Policy 2021-2027. While negotiations in the EU institutions on this regulatory proposal are still ongoing, there is value in reviewing the experiences of stakeholders who have already succeeded in providing CPS without such an instrument in place, so as to better understand the barriers they encountered, how these were overcome, and thereby to contribute to the discussion of how the realisation of CPS might be better supported in future.
Previous analyses have highlighted the considerable diversity in cross-border cooperation across Europe, and the contextual factors, motivations and institutional arrangements that have shaped border region governance in different parts of Europe and at different phases of the European integration process (cf. De Sousa, 2013;Noferini et al., 2020). Arguably this diversity of experiences is nowhere more obvious than along Germany's borders, which include the first Euregio of 1958 on the Dutch-German border, but also much more recent cooperation structures that only became possible following the nation's reunification in 1990 and the subsequent major EU Eastern enlargement in 2004. Recent policy debates have also highlighted the need for tailor-made and bottom-up solutions able to respond to the border region's specific geographical, institutional, political and cultural circumstances, and consequently the need for flexible instruments to support local and regional actors in their efforts to provide CPS (EC, 2021). With nine neighbouring countries and a federal political system, Germany's border regions offer a considerable range of institutional arrangements, and German stakeholders' experiences with the provision of CPS can therefore offer useful insights into the solutions found to overcoming common obstacles.
The paper addresses the following two questions: (1) What types of obstacles did different German border regions experience in providing cross-border SGIs, and how were they overcome?, and (2) based on German border experiences, how could the provision of CPS be better supported?
In order to answer these questions, the following two sections first provide a conceptual discussion on cross-border cooperation and a review of the policy debate on the provision of CPS in Europe. Then, the methodology for the analysis presented in this paper is explained, with the second part of the paper discussing the findings from the empirical analysis of experiences of stakeholders in German border regions in providing CPS. The final section offers conclusions.

Cross-border cooperation in Europe
Cross-border cooperation has been defined as 'any type of concerted action between public and/ or private institutions of the border regions of two (or more) states, driven by geographical, economic, cultural/ identity, political/leadership factors, with the objective of reinforcing the (good) neighbourhood relations, solving common problems or managing jointly resources between communities through any co-operation mechanisms available' (De Sousa, 2013, p. 673).
While traditionally a bottom-up process, voluntary cooperation across national borders was recognised as an important ingredient for the implementation of the Common Market in the 1990s, with the European Commission labelling cross-border regions the 'laboratories for European integration' (Noferini et al., 2020). Reflecting the wider context of European integration, and supported by dedicated policy frameworks of the Council of Europe and the EU, to date four different stages of cross-border cooperation in Europe can be identified (Noferini et al., 2020).
Between 1950 and 1979, cross-border cooperation in Western Europe was mostly conducted on the basis of inter-state agreements, albeit sometimes with supra-national support as in the case of first initiatives of the Council of Europe or the Nordic Council. Since the 1960s, the Council of Europe has been a major promoter of cross-border co-operation as a means to help the diffusion of local democracy and good neighbouring relations (De Sousa, 2013). In this first phase of cross-border cooperation, the first Euroregional initiatives were initiated, following the model of the EUREGIO of 1958 between Gronau (DE) and Enschede (NL).
In a second phase, between 1980 and 1990, cross-border cooperation benefitted from increased European political and juridical support. The 'European Outline Convention on Transfrontier Cooperation between Territorial Communities or Authorities' was initiated by the Council of Europe and came into effect in 1980 (CoE, 1980). Also called the 'Madrid Outline Convention', it provides a legal framework and model agreements for inter-state treaties and local level cooperation, and has to date been signed by 39 countries. The attention to sub-national authorities on the EU's political agenda was also considerably enhanced by the reform of the Cohesion Policy in 1998 during this period.
Between 1990 and 2006 a third phase of considerable political and economic support for cross-border cooperation can be identified, following the launch of the Community Initiative INTERREG in 1990. The adoption of the Maastricht Treaty of 1993 marked the beginning of a phase of deepening, broadening, and (through northern and eastern enlargements) widening of European integration. The Maastricht Treaty was also notable for giving more explicit acknowledgement to the role of sub-national authorities in the EU's multi-level governance system, inter alia through the creation of the Committee of the Regions (Noferini et al., 2020). The establishment of the Single Market programme, supported by regional policy funding, prompted a boom in cross-border cooperation, with Keating (1998, p. 180) suggesting that by the late 1990s 'there was not a border in Western Europe that was not covered by some sort of transfrontier programme'. Following the EU's northern and eastern enlargements, more cross-border cooperation structures were created along the 'new' internal EU borders and also along the new external borders of the enlarged EU (Bufon & Markelj, 2010).
Since 2006, further support for cross-border cooperation has been made available through the EU Regulation on 'European Groupings of Territorial Cooperation' (EGTC) (EU, 2006), adopted as part of the Cohesion Policy 2007-2013reform package, and updated in 2013(EU, 2013. The EGTC Regulation gives sub-national public authorities from different member states the right to institutionalise their cooperation arrangements without requiring approval from their national or state governments to create joint transboundary organisations with their own legal personality (De Sousa, 2013). Since 2007, EU funding for cross-border, transnational and inter-regional cooperation was further institutionalised, through a 'mainstreaming' of the previous Community Initiative INTERREG into the main EU Cohesion Policy objective on European Territorial Cooperation (Dühr, Colomb & Nadin, 2010).
Today, Europe is covered by a considerable number of nested or overlapping transborder cooperation initiatives at different scales (Deas & Lord, 2006;Kaucic & Sohn, 2021), with different legal arrangements and varying scope of institutionalisation and of different intensity of cross-border cooperation. A particular density of cooperation arrangements at different scales can be found in 'Rhineland Europe' and Central Europe (Kaucic & Sohn, 2021). Cooperation arrangements at different scales are usually set up to address specific concerns, such as local-level cross-border cooperation to provide solutions to the daily needs of commuters crossing the border. 'Regional-scale' cooperation usually focuses on 'higher-level' functions (e.g. airports, universities), and cooperation at the supra-regional (macro-regional) scale is often rooted in the search for more efficient management arrangements of major shared ecosystems (regional seas, transnational river basins, mountain ranges) or to coordinate spatial development along major infrastructure connections (Peyrony, 2020;Kaucic & Sohn, 2021). This multi-scalar dimension of cross-border cooperation is important, because as previous research has shown the interrelationships among cooperation arrangements at different scales can be relevant factors in the (co-)evolution and institutionalisation of border regions (Kaucic & Sohn, 2021). Cross-border actors might even strategically use the high-level political attention accorded to macro-level cooperation to address their local-level concerns, to for example improve cross-border transport connections (Dühr & Belof, 2020).
However, the considerable number of cross-border cooperation arrangements across Europe, and at different scales, usually remain add-ons to the work of territorial jurisdictions, rather than challenging the Westphalian model of the nation-state. Much of the previous research on crossborder regions focused on whether more formalised and institutionalised forms of cooperation between local and/or regional governments are prompting processes of reterritorialization and the rescaling of competences from territorial jurisdictions to the 'soft spaces' of border-regional organisations (see e.g. Keating, 1998;Brenner, 2004). Recent border closures in response to the Covid-19 pandemic have shown, however, that despite many years of European integration nation-states remain the central actors for determining international economic relations (see e.g. Medeiros, Guillermo Ramírez, Ocskay & Peyrony, 2021). National and state actors also continue to be responsible for key policy areas relevant for the provision of public services, which can help to explain why in spite of long-standing support by both the Council of Europe and the EU for crossborder cooperation, truly integrated and institutionalised joint projects across national borders remain rare.
In terms of 'local-level' cross-border cooperation, which is the focus of this paper, different motivations for joint working have been identified. De Sousa (2013) offers a useful typology for local-level cross-border cooperation according to the degree of political commitment and institutional intensity. These are not mutually exclusive and may occur in the same border region. They are: (1) Awareness raising co-operation, requiring the lowest level of political commitment (examples are regular bilateral visits or town twinning arrangements to promote cultural and commercial ties); (2) Mutual aid co-operation, i.e. an agreement among emergency responders from both sides of the border to lend assistance across jurisdictional boundaries in case of disasters or other emergencies; and (3) Functional co-operation to address interdependencies across the border (such as shared housing and labour markets) and that usually result in more permanent cooperation and therefore require 'greater resources and a higher degree of commitment from the neighbouring local/regional political and administrative authorities' (De Sousa, 2013, p. 674). Finally, and of relevance to the discussion in this paper, type (4) refers to the common management of public resources/services, and characterises a more limited number of co-operation arrangements that seek 'joint strategies to reorganize and rationalize state services, benefits and other public funded provisions in function of border regions rather than being based on state interests and jurisdictions' (De Sousa, 2013, p. 675).
De Sousa (2013) argues that the institutional basis for the more integrated types 3 and 4 is currently mostly provided by either Euro-regions or EGTCs. Euroregions, and the more urbanised Eurodistricts along the German-French border, are traditionally characterised by a high number of public actors and cooperation between local and regional authorities (Noferini et al., 2020). The ambitions of Euroregions and EGTCs are, however, often hampered by a lack of financial autonomy and because the decisions of their member organisations continue to be subject to political logics primarily focused on the national territory. Moreover, as Princen, Geuijen, Candel, Folgerts and Hooijer (2016, p. 499) have shown for police, fire brigades and emergency health services, cross-border cooperation aimed at the provision of public services may be organised as part of formal cross-border regional cooperation, but there are also numerous cases where cooperation is 'established and operates on the basis of separate agreements or informal understandings between relevant organizations on the two sides of the border'. This suggests that a discussion about CPS requires attention beyond formalised cross-border regional structures and an acknowledgement of different drivers, actor constellations and areas of cooperation, which -depending on the border region -may require specific formal or also informal agreements to ensure their long-lasting provision. In these processes, the importance of 'policy entrepreneurs' or 'regionauts' (Löfgren, 2008) has been recognised. These are 'street-level professionals' or government employees who can place the need for specific cross-border services on political agendas and are able to find creative solutions to overcoming 'differences in legal, organizational and cultural backgrounds between the participating countries' (Princen et al., 2016, p. 498).

Obstacles for cross-border cooperation and the provision of cross-border public services
With the Single European Act of 1986, economic and social cohesion were introduced as key objectives for the then European Community, and aimed at reducing disparities between regions. In 2008, the Treaty on the Functioning of the European Union (Lisbon Treaty; TFEU, 2008) added the objective of territorial cohesion. Economic, social and territorial cohesion are supported through EU Cohesion Policy and its related funding instruments. The objective of territorial cohesion has put the focus on the provision of 'services of general economic interest', also referred to as 'public services' , to the population in all areas of the Union territory.
In the EU policy debate, SGIs are now commonly understood as basic services that are subject to specific public service obligations to ensure certain quality standards, and affordable and equal access for all citizens. Three categories of SGIs have been identified (EC, 2011): • Services of general economic interest (SGEI), which are basic services that are carried out in return for direct payment. They deliver outcomes for the overall public good that would not be supplied (or not be adequately supplied in terms of quality, safety, affordability, equal treatment or universal access) by market providers without public intervention. SGEIs can include technical infrastructures, such as public transport, energy, water and waste; and communication infrastructures, such as postal services and telecommunication services. These services are subject to European internal market and competition rules. However, there may be derogations to these rules if necessary to protect citizens' access to basic services. • Non-economic services (NESGI), such as the police, justice and statutory social security schemes.
These are not subject to specific European legislation or to internal market and competition rules. • Social services of general interest (SSGI) are based on the principles of solidarity and equal access, and can be of an economic or non-economic nature. Examples include compulsory education, healthcare, social security schemes, employment services and social housing. Different EU member states use different definitions and arrangements for the provision of SGIs. The discussion of which service is of 'general interest' is further impeded by the fact that public authorities are in most European countries no longer the sole supplier of such services but rather impose public service obligations on different (public or private) providers of such services. Moreover, Humer and Palma (2013) have shown that EU 15 regions are generally much better equipped in terms of SSGIs such as education and health care, and that the urbanized areas of Western Europe benefit from a higher level of SGEIs. The debate over SGIs therefore reflects the tension over the need for regulating the provision of basic services through private sector providers so that the 'public interest' can be ensured even in regions with low population densities, located in peripheral locations or for border regions where access to SGIs is presenting challenges. Article 174 of the TFEU specifically noted that citizens in border regions are often less well served by SGIs than the population in regions more centrally located in national territories, and that therefore the EU institutions and the member states should pay particular attention to cross-border regions when developing action to strengthen the EU's economic, social and territorial cohesion (van Lierop, 2019).
The concern over the continuing challenges in providing CPS has been at the core of the recent EU policy debate on legal and administrative obstacles to cross-border cooperation. This debate, initiated in 2015 by the member states and complemented by EU analyses, culminated in 2018 in a regulatory proposal by the European Commission for a new legal instrument to help overcome barriers to border integration (EC, 2018; for details on the processes leading to this ECBM proposal see Engl & Evrard, 2020;Rosanò, 2021).
In overview, in the second half of 2015, the Luxembourg EU Council Presidency presented a proposal for an EU legal instrument to overcome administrative and legal barriers to cross-border cooperation that had been prepared with input by the Transfrontier Operational Mission (MOT) (MOT, 2015a(MOT, , 2015bLuxembourg Presidency, 2015a, 2015b. At the Dutch Presidency in 2016, EU member states gave the mandate to establish a 'Working Group on Innovative Solutions to Cross-Border Obstacles' (Engl & Evrard, 2020). The membership of the working group comprised about half of the EU member states and additional stakeholders, with MOT providing the technical secretariat. In 2017, the Working Group presented a proposal for a new legal instrument to address administrative and legal obstacles in cross-border cooperation that would allow one member state to transfer their regulations to the other side of the border for the purposes of the joint initiative (WG, 2017a(WG, , 2017b. In parallel to the work of the member states, the European Commission had between 2015 and 2017 undertaken an EU-wide 'Cross-border Review' to inform the reform of EU Cohesion Policy 2021-2027. The Cross-Border Review included an inventory of 239 legal and administrative obstacles to cross-border cooperation (EC, 2017b). Three categories of obstacles were defined: First, legal barriers caused by the absence of EU legislation in policy fields where a Union competence exists, or by shortcomings in a transposition of Union legislation into national law. Second, legal barriers caused by incoherent or inconsistent domestic laws of EU member states in policy fields where no or only a partial Union competence exists. Third, administrative barriers caused by inadequate procedural and/or adverse behavioural aspects at the local, regional or national levels. The majority of obstacles identified were in the second category, i.e. inconsistencies between national or regional legislation (104 cases), followed by administrative obstacles due to inadequate procedural or administrative practices at local, regional or national level (99 cases). Far fewer obstacles were identified that relate to the existence or lack of EU legislation (EC, 2017b). Consequently, the need for action between member states and within nation-states to effectively address border obstacles has since received considerable attention in the policy debate.
Over 90% of the legal and administrative obstacles identified by the Cross-Border Review (EC, 2017b) were found in five policy areas and their related public services, namely: labour market and education (e.g. mobility of cross-border workers), social security and health (e.g. access to social insurance system; or access to health care services and medical treatment), transport and mobility (e.g. public transport by bus, rail or light rail), industry and trade, and policy planning and public services (emergency and rescue services; public security and police cooperation; and spatial planning and cross-border territorial development planning) (EC, 2017b). A 2017 study commissioned by the European Commission (EC, 2017c) estimated that the removal of just 20% of cross-border obstacles in border regions could increase their GDP by 2%, and consequently that overcoming cross-border obstacles would enable the EU's border regions to take full advantage of their economic potential.
These analyses provided input for a Communication from the Commission in 2017 on 'boosting growth and cohesion in EU border regions' (EC, 2017a). The Communication clearly framed the discussion of cross-border obstacles in relation to the economic performance of border regions and the EU as a whole, and emphasised the costs of the complexity of cross-border cooperation and duplication of service provision, and the resulting economic inefficiencies. The Commission communication (EC, 2017a) set out a ten-point action plan, with point nine making specific reference to the on-going work of the member states' 'Working Group on Innovative Solutions to Cross Border obstacles' and their proposal for a new legal instrument, stating that the Commission would consider ways to take this new instrument forward. The Commission staff working document accompanying the communication on boosting growth and cohesion in EU border regions (EC, 2017d) argued that the proposed legal instrument would have a number of advantages over existing tools, that it would be quicker and more efficient than intergovernmental agreements and also provide stakeholders with greater administrative and legal certainty compared with the bottom-up approach of many initiatives in border regions.
Thus, the proposed ECBM is intended to complement existing legal and financial frameworks for cross-border cooperation, and specifically offer a tool to facilitate the often complex and tailor-made arrangements for the provision of CPS. In terms of existing tools and their identified shortcomings for supporting CPS, the Madrid Outline Convention has been instrumental for strengthening the legal basis for cross-border cooperation by providing a framework for bi-or multi-lateral inter-state treaties (such as the Anholt Agreement of 1991 between the Federal Republic of Germany, the German states of North-Rhine Westphalia and Lower Saxony, and the Kingdom of the Netherlands), on which basis municipalities and regions from both sides of the border can conclude legally binding cross-border cooperation agreements aimed at coordinating decisions. However, for initiatives that affect national standards, such as traffic safety or school-leaving qualifications, state treaties do not offer sufficient detail and legal certainty.
The Regulation on EGTCs (EU, 2006(EU, , 2013 was a response to the need for a stronger legal basis to take established cooperation activities to the next level by providing more institutional, legal and administrative certainty. The EGTC Regulation required member states to adopt national provisions that give sub-national territorial authorities the direct right to establish organisations with their own legal personality. According to Noferini et al. (2020), the main advantages of establishing an EGTC are that it ensures the long-term political commitment of its members; affords greater visibility with respect to third parties; and gives the EGTC the ability to directly enter into contracts and compete for external and European funding. Even so, uptake has initially been hesitant and to this day there is considerable variation across Europe, with most of the currently 80 EGTCs found in the southern and eastern parts of the EU, and none established to date in northern Europe.
The variable uptake might be explained by the perception that establishing an EGTC can present complex institutional challenges and that the advantage to existing cross-border arrangements, such as under the Madrid Convention, are not always evident (CoR, 2015). However, since around 2010 some border regions with long-standing cooperation agreements, such as the 'Eurodistricts' of SaarMoselle or Pamina along the German-French border, have chosen to adopt EGTC status (Peyrony, 2020). This has led some commentators to suggest that EGTCs might in future 'overcome Euro-regions as models of cross-border cooperation ' (De Sousa, 2013, p. 679). Even so, while EGTCs can decide on their own structure and hire staff, their creation does not result in a transfer of competences from existing authorities, and they therefore cannot offer comprehensive solutions for the provision of CPS. Using the example of the EGTC for the Cerdanya hospital, Peyrony (2020) explains that the French and Catalan health systems retain their competences, with the mission of the EGTC restricted to coordinating these national rules and funding mechanisms.
Finally, while the EU has provided considerable financial support for cross-border cooperation since 1990, this funding has rarely led to the provision of CPS. The EU's 'KEEP' database, which provides an overview of EU-funded territorial cooperation projects since 2007, shows that most cross-border INTERREG projects continue to focus on 'softer' measures such as joint events or feasibility studies that do not require longer-term solutions to be found for administrative or legal differences. There are likely many reasons for the often intangible outcomes of INTERREG projects, including the since the 2007 reform increasingly complex administrative requirements of INTERREG programmes and projects. These have been argued to present barriers for many, especially smaller, organisations, and may even add challenges to finding solutions for complex 'durable' joint projects given the regulatory requirements and the time constraints of the Cohesion Policy funding framework (Bachtler & Polverari, 2007;Dühr et al., 2010).
The policy initiatives of the member states' working group (WG, 2017a(WG, , 2017b and of the European Commission (EC 2017a, 2017b) were eventually brought together in a proposal by the European Commission for a regulation on a mechanism to resolve legal and administrative obstacles in a cross-border context (ECBM) (EC, 2018). This was presented as part of the May 2018 legislative package for EU Cohesion Policy 2021-2027. The proposed regulation stipulated that the ECBM would be a voluntary instrument that would enable one member state to apply the laws of a neighbouring member state for the purposes of a common cross-border project, if legal differences would otherwise present obstacles to the delivery of the joint project.
The proposal for the ECBM Regulation was welcomed by stakeholders, such as MOT, the Association of European Border Regions (AEBR), the European Parliament and the Committee of the Regions. However, some member states -including Germany -raised critical questions in Council discussions about the voluntary nature of the proposed instrument, extra administrative burdens, and its implications for the principles of subsidiarity and proportionality (van Lierop, 2019;Engl & Evrard, 2020;Rosanò, 2021). As a result, the regulation has not yet been approved as part of the Cohesion Policy reform package, but a recent report by the European Commission on border regions as 'living labs of European integration' confirmed that the Commission remains supportive of the proposal and that 'the ECBM is still being discussed in Council, after the European Parliament established its broadly positive position in 2019' (EC, 2021, p. 3).
Yet, while the EU policy debate has in recent years almost exclusively focused on the challenges presented by legal and administrative obstacles, this paper argues that a wider perspective on the barriers to the provision of CPS might be useful, especially considering the heterogeneity in arrangements for the provision of SGIs across the EU. A major survey on obstacles and solutions to cross-border cooperation, undertaken for the Council of Europe in 2013 (ISIG & CoE, 2013), is helpful in this endeavour of providing a more comprehensive insight into the challenges experienced in border regions in realising joint public services. The study proposed a typology of border obstacles based on criteria such as the scale required for a solution (local, national or European); the policy sector concerned (employment, transport, etc.); and the nature of obstacles (ISIG & CoE, 2013, p. 77). Drawing on this study, the following categorization of obstacles will provide a framework for discussing the findings from the empirical analysis in the remainder of this paper: 1. Institutional and legal obstacles (incl. availability of relevant legal frameworks) 2. Administrative obstacles (incl. degree of state centralization; involvement of actors with relevant competences for cross-border cooperation) 3. Economic obstacles (incl. level of socio-economic development on both sides of the border; fiscal barriers) 4. Obstacles linked to the level of expertise of stakeholders involved (incl. language skills) 5. Cultural and political obstacles (e.g. political agendas of key stakeholders; lack of trust; history of cooperation) 6. Obstacles linked to the degree of propensity to cooperate of stakeholders involved (incl. availability of funding).

Methodology
The methodology for the analysis presented in this paper draws on a comprehensive review of the scholarly and policy literature on cross-border cooperation and the provision of cross-border SGIs. This desk analysis was complemented by primary data collected through interviews and a focus group workshop with representatives of German border regions. Germany offers a unique insight into different cross-border cooperation experiences, given that its nine neighbouring countries include 'old(er)' EU member states (Netherlands, Belgium, Luxembourg, France, Austria, Denmark), 'new(er)' EU member states (Poland, Czechia), and one non-EU member state (Switzerland). This, together with the federal system of Germany that provides the German Länder with competences for some of the policy areas relevant to the provision of SGIs, results in a range of different frameworks that shape the approaches to providing CPS. However, identifying cases of institutionalised public services across Germany's borders proved more challenging than expected. In spite of the partly long-standing cooperation arrangements and considerable range of cross-border activities with participation of German actors, truly institutionalised cross-border initiatives are still rare. No database on CPS existed when the research was conducted, although a recent ESPON study (2020) has since started to inventorise practice examples. To identify suitable cross-border initiatives, an extensive internet search, including the review of different border regions' websites, was therefore conducted. This, together with wordof-mouth referrals, led to the identification of a range of examples of different SGIs from Germany's border regions. The following Table 1 provides an overview of the altogether ten examples of SGEIs, SSGIs and NESGIs discussed with representatives from the respective border regions during the workshop. Cross-border public transport is clearly an important concern for German border regions and their commuters, but SGIs have also been realised in other sectors, as Table 1 shows.

Germany-Belgium-Netherlands border
The workshop with representatives from these different German border regions took place in January 2018. Detailed semi-structured phone conversations with the representatives were conducted prior to the workshop to gain an understanding of the different cases of CPS provision. The public authorities and private companies interviewed and invited to the workshop were from the German border regions (rather than also participants from the respective neighbouring country), because the discussion also covered the particularities of the German federal and multi-level system in shaping cross-border initiatives. Other workshop participants, besides the research team, were representatives from German federal departments and agencies.
During the first part of the workshop, each participant gave an overview of their key initiatives for the provision of cross-border SGIs in their region, and discussed the most pertinent obstacles to their realisation as well as how these were eventually overcome. The second part of the workshop focused on discussing approaches that the participants had experienced as helpful in finding solutions to overcoming the main barriers to the provision of CPS. Participants were further invited to reflect on instruments or approaches, including the proposed ECBM regulation, that could facilitate joint SGI initiatives in future.
The analysis presented in this paper can of course not provide a complete and comprehensive overview of the challenges experienced in the provision of CPS in Germany's border regions. Rather, the intention of this paper, given the dynamic policy context and the challenges inherent in identifying existing cross-border SGIs, is to provide a qualitative insight into the work undertaken by 'regionauts' and their organisations in trying to ensure the provision of public services across national borders, and to thereby provide input into the academic and policy discussions on cross-border cooperation beyond the current focus on legal and administrative obstacles.

German stakeholders' experiences with obstacles and solutions for realising cross-border public services
In this section, the main findings in response to the first question ('What types of obstacles did different German border regions experience in providing cross-border SGIs, and how were they overcome?') are discussed, using the six types of obstacles defined above to structure the discussion. Generally speaking, the analysis confirmed previous accounts of the need for tailor-made solutions, and all participants noted that usually a considerable range of legal and administrative arrangements is needed for cross-border SGIs. Office of Transport, the canton of Basel-Stadt, Basler Verkehrsbetriebe and the City of Weil am Rhein (Germany); and a planning, construction, operation and financing agreement between the two transport operators (Basler-Verkehrsbetriebe and Stadtwerke Weil am Rhein). For one of the SSGIs under study, the German-Luxembourg Schengen Lyceum in Perl, an agreement between the Grand Duchy of Luxembourg, the Saarland Parliament and the Federal Republic of Germany on the basis of the inter-state treaty and by ministerial decision was required. This agreement regulates the location of the joint school and its name, the school authority and legal status, and objectives for the curriculum. Furthermore, an administrative agreement on staffing and pay scales, school supervision, holiday regulations, etc., and a financial protocol were required.
Among the major institutional and legal obstacles experienced, German stakeholders noted that there are usually a number of legal obstacles to be overcome in the provision of CPS, not least because of different technical standards (e.g. electrification and safety standards for public transport, registration of vehicles) or differences in quality standards (e.g. drinking water). This detail and complexity means that it usually takes several years from initial memorandum of understanding to practical realization, and trial-and-error was often needed before 'workable' arrangements could be found. For example, for the joint bus route Frankfurt/Oder-Słubice the initial ambition was to set up a formal agreement on the joint operation and financing of the bus. This was not feasible, and in the end it was agreed that the tariff area of the Berlin-Brandenburg transport association (VBB) would be extended to include Słubice, which was formalised through the direct commissioning of Stadtverkehrsgesellschaft mbH Frankfurt/Oder by the municipality of Słubice for the section of the route on the Polish side. Resolving all issues arising from differing technical standards and complex pricing and ticketing issues for the extension of the Strasbourg tram line in France across the border to neighbouring Kehl in Germany took almost ten years and some compromises (EC, 2017b). Eventually, it was agreed to use the French trams for the cross-border route, and to retrofit them with additional reflectors to ensure compatibility with German requirements. Agreements were also needed for signalisation for road and tram traffic and on ticketing, pricing and mutual recognition of travel tickets (WG, 2017b). Although the actual planning and construction of the tramline was the responsibility of the involved French and German local authorities, the project was thus strongly shaped by national regulations and standards for traffic safety. Moreover, and as identified by the Commission's Cross-Border Review (EC, 2017b), in addition to institutional and legal obstacles arising from national differences there are also obstacles in areas covered by EU law. For example, approaches to reaching water quality standards as required by the EU Water Framework Directive were interpreted differently in the respective national laws in Germany and France, and this presented considerable challenges for reaching agreement on the joint drinking water supply project. The workshop participants also reported on initiatives that could not be realised to date as no solutions to legal differences could be found. For example, a joint horticulture exhibition between Strasbourg and Kehl could not yet be realised because French local authorities are not allowed to be involved in private enterprises (such as a limited liability company).
In terms of administrative obstacles, the workshop participants noted that differences in the distribution of competences between levels of government, which are especially pronounced between centralised and federal systems, can present considerable barriers in realising CPS. Greater commonality in the division of competences between both sides of a border was perceived as facilitating joint initiatives. For example, according to AOK Bayern it was easier to reach cooperation agreements on cross-border health care with Austria and South Tyrol, i.e. federal or regionalised nations, whereas considerably greater difficulties were experienced in discussions with health care providers in the centralised Czechia. For the Schengen Lyceum, originally also French participation was envisaged, but could not be realised because the country's centralised school system proved prohibitive for finding decentralised flexible solutions for the joint school.
In terms of economic obstacles, the lack of a clear legal basis for joint tendering and financing was mentioned as an obstacle to the implementation or continuation of joint projects. For example, for the cross-border bus line 983 between Frankfurt/Oder and Słubice, no permanent agreements that would include the transfer of competences between the involved authorities could be achieved to date, and differences in municipal budgets and rules for accounting have been reported as presenting challenges for the long-term provision of the bus connection. Moreover, several participants noted that fiscal differences (e.g. in VAT) between nation-states can pose problems in awarding tenders and implementing joint projects.
In terms of obstacles related to the level of expertise of cross-border stakeholders, the discussions in the workshop highlighted the crucial role of 'policy entrepreneurs' in realising complex CPS. In the absence of standardised routines, cross-border cooperation is highly dependent on the motivation of individuals and the support of their organisations, with personal relationships and trust between actors essential for achieving tangible outcomes. The success of CPS can thus often be ascribed to the commitment of a small number of policy entrepreneurs from the different sides of the border, who are willing to conduct often complex negotiations over many years and who are able to find solutions even in the face of changing framework conditions. This requires persistence and continuing engagement with those actors, including politicians, who are not so closely involved in the daily realities of cross-border work. This is because even after decades of cross-border cooperation it cannot be assumed that mutual understanding has been reached across all levels of government and administration. One example given was that Franco-German agreements are often understood differently on both sides of the border, because of different interpretation of what is lawful (in the words of one of the participants: 'In Germany, what is not forbidden is allowed, in France, only what is explicitly allowed is actually allowed'), and that such differences frequently (and repeatedly) require explanation by experienced 'regionauts'.
While legal and administrative obstacles have received most attention in recent policy debates, workshop participants emphasised that political and cultural factors can play a very important role in realising CPS. Among the border regions represented at the workshop there were different histories of cooperation, notably between Germany's Western and Eastern borders. Some regions that are characterised by long-standing cooperation along Germany's Western borders have over time succeeded in realising a range of joint public services to address cross-border interdependencies and set up functional cooperation. For example, the Strasbourg-Kehl border region has set up a Franco-German fire brigade, a joint crèche and a joint tram line. However, independent of the history of cooperation, political support for cross-border initiatives, both at local and regional levels as well as -depending on the requirements of the project and the division of responsibilities within the countries involved -from national governments, was by all participants considered to be absolutely essential for realising complex SGIs. However, the participating project initiators noted that crossborder work is still -and after many years of institutionalised cooperation -considered by many government departments and politicians to be distant from their everyday concerns and that awareness of the challenges involved in realising joint services through often tailor-made solutions by decision-makers cannot necessarily be assumed. Indeed, some participants even argued that cross-border cooperation today is more challenging than it was during the 1990s when pro-European attitudes prevailed. Participants felt that large pro-European 'special solutions' of the 1990s (such as the joint business park Eurode between Germany and the Netherlands) would hardly be possible today. This is not only a challenge for cooperation with Germany's neighbouring states that have seen a shift to populist governments in recent years, such as Poland, and where it has since become more difficult to hold regular cross-border exchanges. Rather, it was argued that cross-border cooperation in the current EU context is strongly influenced by changing political priorities at all levels, and that this may even jeopardise already formalised agreements. For example, a change in the government of Strasbourg resulted in a reconsideration of all proposed cooperation projects, causing delays of several years. Consequently, workshop participants felt that increasing awareness of the need for cross-border approaches and their complexity is essential, and that there needs to be greater political awareness and political support for the flexibility required to realise CPS.
The importance of political support for cross-border cooperation can thus not be overstated. Indeed, workshop participants reported that if there is political support on both sides of the border, there is usually flexibility to apply the more favourable national law to achieve the common goal. For example, for the Strasbourg-Kehl border area it was reported that for urban design competitions German law offers more possibilities than French law and is therefore applied to such processes. On the other hand, French law allows for the tendering of planning and construction with a cost cap and with risks to be borne by the company undertaking the work, which makes these rules more attractive for joint tendering processes. However, even with political support and experienced 'regionauts' driving the initiative, all participants noted that finding common solutions to the provision of CPS is time-consuming and requires creativity and perseverance. Joint working groups involving representatives of relevant authorities from both sides of the border are usually set up to manage the process towards finding a solution, requiring commitment over several years. For example, it was reported that the Schengen Lyceum working group held about 100 meetings over a period of two years.
Funding can present considerable obstacles for the realisation of joint SGIs, and influence the propensity of stakeholders to cooperate. Most of the examples discussed during the workshop required complex financing arrangements, in addition to legal and administrative agreements. For example, the Tram 8 between Basel and Weil am Rhein was co-financed by the Swiss Confederation, Canton of Basel-Stadt, Basler Verkehrsbetriebe (transport authority), Industrielle Werke Basel (regional employer), the German State of Baden-Württemberg, the City of Weil am Rhein and the District of Lörrach, as well as INTERREG funding. While EU funding through INTERREG was considered important to provide incentives to cooperate, the coordination of national funding guidelines with EU funding requirements was described by participants as complicated and time-consuming. Moreover, especially since the reform of Cohesion Policy in 2007, the administrative burden of applying for and implementing INTERREG projects was considered by many participants to be unreasonably high and therefore increasingly less attractive. This administrative burden was noted especially for the (in comparison to less developed regions) modest co-financing rates along Germany's western borders.

How could the provision of cross-border public services be better supported? Views from German border regions
In order to answer the second research question, workshop participants were invited to reflect on approaches and instruments -including the then ongoing work on the ECBM -that could support the realisation of CPS. Although all border regions involved in the study were covered by inter-state treaties and/or EGTCs, participants acknowledged that these frameworks are not sufficient to arrange the details of CPS provision, and that additional agreements were necessary for all of the examples provided to ensure lasting arrangements. Many participants felt that the EGTC regulation was overly complex, and that especially in regions with long-standing cooperation the advantages to existing inter-state treaties based on the Madrid Outline Convention were not always apparent.
What the workshop discussions clearly showed is that every realised CPS requires considerable commitment, flexibility and persistence. In finding solutions to the obstacles, cross-border policy entrepreneurs may test and apply a range of options. For example, the representative from Kehl suggested that in their border region currently three models for the realisation of CPS are used. First, a 'transfer model', whereby an approach that has been used in one country is transferred to the other country. The example given is the French substitution practice in drug counselling, which has informed practice in Kehl since 2013 (Egg, 2016). Second, an 'adapter model', whereby a custom-designed 'adapter solution' reconciles different national models in a joint project, such as applied to the hose coupling of the German-French fire brigade. Finally, the 'supreme model' includes the development of new arrangements that are tailored to the needs of the border area and reconciles national regulations and standards. Examples from the Strasbourg-Kehl border region are the joint fire-fighting boat and the cross-border crèche. Political support is important to realise any CPS, but is absolutely key for more integrated solutions as needed for a 'supreme model' or 'adapter model'.
Workshop participants called for more consideration of the special circumstances for crossborder service provision, and argued that a more flexible implementation of national rules in border areas or adapter solutions should be permitted if there is a clear need for the joint initiative, even if they require derogation from national legislation. This obviously requires the ability, and competence of decision-makers, to consider the value of alternative solutions ('other rules can also make sense'), and scope to translate this into pragmatic arrangements, such as allowing national funding to be spent on the other side of the border. In this vein, workshop participants argued for more policy attention to intercultural communication and the education of politicians and others involved in cross-border cooperation. EU law but also bilateral agreements are often interpreted very differently on the two sides of the border, which requires constant dialogue and clarifications.
In order to support the realisation of CPS, workshop participants emphasised the need for more coordinated support at all government levels able to inform and guide the work across borders. This covered a number of suggestions, including the establishment of a 'clearing house' at EU level and/or 'legal tandems' within the member states tasked with facilitating the implementation of complex border area projects. These bodies would be able to advise on the specific legal implications of proposed initiatives from an EU perspective or respectively from both sides of the border, thus providing a 'one-stop-shop' for all legal questions arising for a specific initiative. This could be complemented by cross-border regulatory impact assessment procedures, considered promising tools to improve the harmonisation of EU law between jurisdictions and to 'test' the cross-border compatibility of national regulations, such as in the case of planning for wind turbines where differences in standards about distances to other uses can lead to conflicts over windfarm developments near national borders.
There was considerable discussion about the specific challenges for local authorities in border regions, given that they are key stakeholders for 'local-level' cross-border cooperation, yet despite the rhetoric of being the relevant scale for realising practical responses to European integration are often not sufficiently well equipped in terms of autonomy, competences and also funding. For example, workshop participants emphasised that a more comprehensive consideration of the implications of the freedoms of the EU Single Market on border regions would be important to support high-level goals of social and territorial cohesion. Labour markets and housing markets are increasingly interdependent in many border areas, but housing policies are not usually coordinated across borders and the impacts of developments on municipal finances are rarely systematically considered. In regions with considerable imbalances in housing supply and/or affordability and employment opportunities on different sides of the border, this can put considerable strain on municipal facilities such as kindergartens or swimming pools, the demand for which is not usually calculated with cross-border interdependencies in mind but rather based on national formulas.
More generally, workshop participants argued that there needs to be more recognition by the EU, national and regional governments about the comparatively higher financial burden on municipalities in border regions to provide SGIs. Expensive 'technical' infrastructures for SEIGs such as water treatment plants or public transport connections require considerable legal certainty and 'protection' against changes in national legal or political framework conditions to enable their operation over many years and to give municipalities longer-term certainty over their financial planning. Moreover, participants felt that this recognition of the often much more complex requirements for cross-border initiatives than for 'domestic' projects should be reflected in more financial support from the EU and from national (and state or regional) governments. In this context, workshop participants called for a refocusing of INTERREG funding on the original intention of addressing specific border obstacles, rather than having the funding directed to the EU's generic jobs and growth objectives.
Given the evident and continuing complexity of CPS provision, what was the workshop participants' view on the proposal for the ECBM Regulation? While the Commission's proposal had not been released at the time the focus group workshop was held, the proposals of the intergovernmental working group for a new legal instrument provided sufficient information for a reflection on the usefulness of the proposed legal instrument by representatives from Germany's border regions. The proposal was in principle welcomed as a possibly useful addition to the toolbox for more institutionalised cross-border cooperation. Defining the area of the proposed initiative as 'experimental space' could facilitate agreements on special solutions for joint public services, such as police cooperation or tariffs for cross-border public transport. More generally, workshop participants felt that the EU-wide debate on the ECBM proposal could potentially contribute to raising awareness for the specific circumstances of border regions outside the small group of 'believers'.
However, while the principle of the proposal was welcomed, participants raised questions about the practical application. Given the considerable diversity of policy areas regulating public services and the range of institutional and political frameworks in the EU's border regions, the question was raised how this instrument could be sufficiently flexible to the needs of different border regions and the variety of proposed cross-border projects while not also resulting in additional administrative burdens. Moreover, participants emphasised that legal and administrative barriers are only one of the challenges of realising CPS. Political and cultural barriers were perceived to be at least as, if not more, important in finding solutions. Participants pointed out that if there is political support for the project, and for cross-border cooperation more generally, then the application of the 'more appropriate' national law is often already possible without an instrument such as the ECBM in place. At the same time, workshop participants felt that the proposed regulation might offer stability from political changes, which could be particularly useful in border regions where political support for cooperation can be variable.

Conclusions
This paper reported on the experiences of German stakeholders with realising CPS. Two questions guided the discussion presented in this paper. In response to the first question, what types of obstacles have been experienced in German border regions in realising joint SGIs and how these were overcome, it is clear that institutional, legal and administrative differences indeed present major barriers, and considerable time, experience and commitment is usually required to find solutions for the provision of CPS. However, the current policy focus on legal and administrative obstacles rather distracts from the importance of other obstacles. Notably, local and high-level political support; financial incentives to address complex border challenges; and dedicated local and regional 'policy entrepreneurs' were considered by workshop participants to be at least as important for realising complex joint projects. Indeed, workshop participants emphasised that if there is political support for the project on both sides of the border, then a solution can usually be found even for involved and challenging legal and administrative obstacles.
This also means, however, that changes in national political priorities can have lasting effects on cross-border cooperation and even call into question already existing legal foundations for joint facilities. For example, recent years have seen the discontinuation of Danish-German cooperation for joint air ambulance rescue services, with austerity measures in Denmark cited as the reason (Jung & Jesumann, 2014). Local elections can also have a lasting influence on the political support for CPS, and even ongoing projects can be jeopardised by the time-intensity and hurdles experienced in realising complex cross-border projects, especially if they are faced with changeable political support.
In response to the second question of how, according to German border stakeholders, the provision of CPS could be better supported, the proposed ECBM Regulation was considered a potentially useful addition to the existing set of instruments for cross-border cooperation in Europe. It was seen as a practical and purposeful means of enabling 'adapter' solutions for many of the legal and administrative barriers experienced in realising CPS. The visibility of a new EU legal instrument was also considered to potentially help with raising awareness among a wider circle of actors for the need to find solutions. However, respondents felt that such a new legal instrument would not be able to address all challenges, notably those identified in relation to political and cultural factors, expertise and propensity to cooperate. Additional tools, and a review of existing instruments and approaches, would therefore be needed to facilitate the realisation of CPS. Workshop participants suggested the need for a critical review of funding regimes such as INTERREG A to ensure they offer real incentives to tackle the considerable challenges inherent in realising cross-border SGIs. Moreover, impact assessments on the potential impacts of proposed legislative changes in the member states and of new EU legislation on cross-border cooperation would be important tools to support the provision of CPS. The recent EU 'Better Regulation' package that proposes territorial impact assessments and 'border-proofing tests' for new legislative proposals (EC, 2021) might -once more widely applied -address such concerns. Finally, the important role of 'policy entrepreneurs', that is, professionals employed by the public authorities or service provider agencies, could be better recognised and their efforts be supported through additional capacities, given that the realisation of a CPS requires considerable creativity and commitment, often over many years, before a solution to the challenges presented by national frameworks can be found.
In conclusion, cross-border cooperation is an essential part of the European integration project, but it is vulnerable to pan-European and global developments. At a time when Europeans are faced with a series of crises that have contributed to a rise of populist politics and Euroscepticism in many parts of Europe, and with border closures in response to refugee flows, terror attacks and the Covid-19 pandemic most acutely felt in Europe's border regions, workshop participants felt that now more than ever is there a need for proactive political education that emphasises the value of open borders and of integrated border communities that are supported by efficient joint public services. Proposals for new legal instruments able to lift cross-border cooperation into more 'mainstream' policy and implementation arenas, rather than continuing to rely on individual projects and tailor-made solutions for CPS, should therefore be welcomed, but they will need to be supported by other measures aimed at increasing political awareness and providing sub-national border authorities with sufficient capacity and financial resources.