Soft planning in macro-regions and megaregions: creating toothless spatial imaginaries or new forces for change?

ABSTRACT Both planning practice and research increasingly acknowledge the existence of new scales and governance arrangements alongside and between statutory planning systems. Examples of new scales of non-statutory planning are large-scale megaregions and macro-regions. Drawing on examples from North America and Europe (Southern California and the Danube Region respectively), this article examines how new processes of cooperation at this scale can influence other statutory levels of decision-making on spatial development. The analysis of spatial delineations, discourses, actors, rules and resources associated with megaregions and macro-regions suggests that this type of ‘soft planning’ can foster territorial integration when a perception exists that there are joint gains to be made, when informal rules are negotiated in context-specific and bottom-up processes, when soft spaces are used as arenas of deliberation to renegotiate shared agendas, and when actors succeed in ensuring the anchorage of informal cooperation in other arenas.


Introduction
Planning practice and research is witnessing the emergence of new scales and governance arrangements both alongside and between statutory planning systems, a feature that has been described in terms of 'soft spaces' (Allmendinger and Haughton 2009;Purkarthofer and Granqvist 2021). Soft spaces can be found in various territories and scales, ranging from sub-local to supra-national . The EU macro-regions and US megaregions discussed in this article are examples of large-scale, non-statutory soft spaces. In Europe, macro-regions are cooperation arrangements between nation states to identify joint development pathways, often linked to spatial development visions and further territorial integration. Meanwhile, US megaregions employ a range of different governance arrangements and cooperation approaches to coordinate regional development at the large scale.
The emergence of these new scales and governance arrangements raises questions about the extent to which they influence processes and decisions of statutory spatial planning. By critically examining the Southern California megaregion (So-Cal) in North America and the Danube macro-region (EUSDR) in Europe, this article sheds light on how these soft spaces not only represent an additional scale in the realm of spatial planning, but facilitate spatial policy development and delivery through new governance arrangements. The article thus investigates the relationship between these new scales and subsequent actions, and explores whether the notion of collaborative, non-statutory 'soft planning' (Faludi 2010) can be used to explain these processes. By doing so, the article takes up a new analytical perspective compared to earlier studies which often focus on the creation of soft spaces and the softening/hardening of territorial boundaries and organizational entities (Metzger and Schmitt 2012;Zimmerbauer and Paasi 2020).
According to Haughton et al. (2010), soft spaces are elements of a new form of spatial planning, where the spatial reach of planning processes finds fuzzy boundaries and covers different sets of issues and actors than in statutory spatial planning. Especially in the context of EU policy making, these trends have also been conceptualized as 'soft planning', referring to processes of strategy development, coordination, cooperation, negotiation and learning (Purkarthofer and Granqvist 2021). While Faludi (2010) proposed that soft spaces require soft planning, Zimmerbauer and Paasi (2020) identify an unclear relationship between the two concepts, claiming that statutory planning can also produce soft spaces, and that soft planning can take place within formal territorial spaces.
EU macro-regions have been discussed as examples of soft spaces and soft planning, referring to their informal institutional structures and the flexibility regarding participating actors and activities (Allmendinger, Chilla, and Sielker 2014;Faludi 2010;Gänzle et al. 2018;Stead 2011Stead , 2014. Unlike macro-regions, US megaregions have not been conceptualized as soft spaces, but they have not gone unnoticed in the academic literature. For example, Innes, Booher, and Di Vittorio (2011) have described mega-regional strategies as mechanisms for collaboration and network building among diverse actors without legislative or bureaucratic authority, even though they are often advocated and encouraged by governmental actors. Meanwhile, Schafran (2014) considers megaregions as means of intervention through alliances across governmental and non-governmental lines rooted in the multiple smaller and 'more real' spaces within the megaregion. These descriptions highlight some of the same characteristics that led to macro-regions being considered as soft spaces.
This article aims to extend the conceptualization of US megaregions in the context of soft spaces and soft planning. This resonates with the claim that the literature on soft spaces and fuzzy boundaries offers interesting avenues worth exploring for conceptualizing multijurisdictional governance in the US (Brown and Shucksmith 2017). The paper also responds to observations that, rather than identifying megaregions, the challenge is to recognize their position, role and status in a broader context (Harrison and Hoyler 2015). Thus, the paper contributes to clarify what 'megaregional planning' means in more specific and practical terms (Harrison and Gu 2021).
While there is little doubt that new scales and governance arrangements have gained importance vis-à-vis statutory planning processes, the emergence of soft spaces and soft planning has also raised a number of political concerns. Olesen (2012), for example, claims that soft spaces serve as vehicles for neoliberal transformations by promoting policy agendas centred on economic development and prioritizing certain interests and policies over others. Soft spaces have also been criticized for their limitations regarding participation, legitimacy and democratic representation (Allmendinger, Chilla, and Sielker 2014). In particular, the increasing complexity of policy-and decision-making can make it unclear where planning can be found for those who want to engage in the process . As soft planning does not follow universal or transparent rules, it is not always clear who has the right to participate, make proposals and decide when it comes to soft planning, let alone how plans are implemented and who has to abide them (Purkarthofer 2016).
Despite the critical debate on soft spaces related to aspects of neoliberalism and democracy, it is rarely questioned whether there is a 'need' for soft spaces, or whether soft spaces in fact facilitate policy delivery and contribute to resolve challenges, as they are frequently claimed to do (Allmendinger and Haughton 2009). This is especially questionable in the context of large-scale cooperation spaces and processes, which do not always show immediate connections to local or regional planning processes and are often not relatable for citizens, public servants and politicians (Sielker 2016b(Sielker , 2017. To address this research gap, this article poses the question how soft planning in large-scale soft spaces influences spatial planning and governance practices at other territorial scales.
The main contributions of this article are threefold. First, it offers a systematic analysis of two examples of soft planning, contributing to a better understanding of the characteristics and limitations of soft planning. Second, it takes a comparative perspective in discussing the specificities, commonalities and potentials of large-scale cooperation processes in Europe and North America. Third, the article analyses not only the 'how' of soft planning but also questions the 'why', considering how these processes can have an added value from the perspective of planning and spatial development.
2. New forms of governance and planning: soft planning for soft spaces 2.1. Development of non-statutory spaces and scales Territory and scale have been subject to theory-building for several decades, sparked by the growth in critical thinking in social theory often affiliated with neo-Marxism starting in the 1970s (Raffestin 2012;Soja 1971). Since the 1990s, scholarship in the social sciences has addressed different conceptualisations of space and territoriality, often in response to changes triggered by processes of globalization and re-regionalisation (Cox 2013;Haughton et al. 2010). Initial conceptualisations of space were rooted in the Westphalian understanding of the nation state as sovereign power over a 'fixed territory', delineated through political negotiations or conflicts. Taylor (1994) described this static understanding of the state as 'container' space. Contrasting the container view, Castells (1989) referred to 'spaces of flows', describing how spatial relations are determined by flows instead of linear state borders in a 'network society'. This resulted in a shift towards processual, dynamic conceptualisations of space in the early 2000s, highlighting the social, porous and networked nature of spaces (Amin 2004;MacLeod 2001;Massey 2005).
These notions of constructivism and relational understandings of space, guided by post-structural and political-economic academic debates, not only questioned the spatial delineation of the state but also its functioning, reflected for instance in Brenner's (2004) argument about the 'hollowing out' of the nation state due to the increased influence of the international and regional level. The state-centred territorial view was thus increasingly superseded by the emergence of a regional geography (Davoudi 2012;MacLeod and Jones 2007), entailing a plurality of interpretations as to how regions are to be defined or characterized (Paasi 2012;Paasi and Metzger 2017). Recent contributions have aimed to overcome the binary distinction between territorial and relational space and to acknowledge both the boundedness and porousness of regions (Harrison 2013;Jones and Paasi 2013;McCann and Ward 2010).
The changing approaches towards space and state mirror the developments in the conceptualization of scale. Challenging the concept of 'fixed' and 'bounded' scales led to a procedural understanding of scale (Marston, Jones, and Woodward 2005) and the acknowledgement of multiple layers of interventions and interactions (Goodwin 2013). One outcome of rescaling is the creation of new spaces or territories in which policy-making or programming occurs. However, the functional, political and institutional boundaries are not always coincident, potentially resulting in a mismatch between issues and impacts on the one hand and resources and powers on the other (Allmendinger, Chilla, and Sielker 2014;Faludi 2018). Moreover, shifts in the influence and competences of different levels of decision-making have not always been accompanied by corresponding shifts in resource allocation (OECD 2001).
New, non-statutory scales between and alongside statutory planning systems have been conceptualized as 'soft spaces', acknowledging the existence of overlapping spaces with fuzzy boundaries that lead to new spatial representations and areas of joint intervention (Allmendinger and Haughton 2009). Spatial imaginaries, referring to 'mental maps' which communicate ideas of new spatial delineations through discursive tactics and material practices play a decisive role in the creation of such soft spaces . Soft spaces recognize the co-existence of both fixed and relational spaces and the need for planning to operate through both clearly set boundaries for formal plans and complex relational geographies. Planning studies in particular have welcomed this last conceptual shift, as planning still operates in bounded spaces and searches for a 'legal fix', while constantly aiming to identify the best perimeter to address development challenges of increasingly interconnected spaces. Since the introduction of this terminology, the concept of soft spaces has been applied to numerous case studies, many of which are located in North-West Europe (Purkarthofer and Granqvist 2021).

On the relationship between soft spaces and planning
Soon after the concept of soft spaces was first used, Faludi (2010) referred to the related notion of 'soft planning', arguing that 'soft spaces require, not hard planning that invokes statutory powers […] but soft planning that relies on a joint formulation of strategy, while retaining dispersed, and thus flexible, powers of action' (21). Many processes framed by European Union activities have since been labelled as soft planning (Purkarthofer and Granqvist 2021), including European Territorial Cooperation initiatives (Purkarthofer 2016;Stead 2011Stead , 2014. In this context, the term soft planning does not only denote planning processes in soft spaces, but potentially also implies changing characteristics of planning. Soft planning brings about new forms of cooperation, coordination, negotiation and learning between different public and private actors (Purkarthofer and Granqvist 2021) and can thus be understood as collaborative process in response to a lack of legal frameworks and established ways of doing things. However, soft planning often remains tied to formal governance structures, mainly through the involvement of actors who derive their authority from spaces of bounded territoriality such as municipalities or nation states. Through these connections, soft planning can influence decision-making at other 'hard' spatial scales, both explicitly through coordinated action or implicitly through the reshaping of goals, values and knowledge (Gløersen et al. 2017;Purkarthofer 2018). Despite the growing body of literature on soft spaces and soft planning, Zimmerbauer and Paasi (2020) claim that the relationship between the two concepts remains unclear, particularly whether one implies the other.
In this article, we investigate non-statutory cooperation and governance arrangements at the mega-/macro-regional scale from the perspective of soft planning. By looking at two very different examples of transboundary soft spaces in the US and the EU, we aim to address the question to what extend these cooperation processes represent new forms of soft planning, and what we can learn from these cases about the characteristics and limitations of soft planning.

Analytical framework and methodology
This section presents the framework used to analyse the So-Cal megaregion and the EUSDR macroregion. While the terms megaregion and macro-region prevail in the US and EU policy discourse, the same terms are also used elsewhere (e.g. Harrison and Gu 2021). Moreover, there are some fundamental differences between US megaregions and EU macro-regions, not only in terminology but also in delineation, mandates and rationales, which are highlighted in the following sections. It is also worth noting that some European megaregions that are unrelated to the EU macro-regional strategies have been discussed previously (see for example Grundel 2021).
Despite their differences, So-Cal and the EUSDR do share some common characteristics from the viewpoint of spatial planning. While they are not part of the statutory planning system and have no direct or hierarchical influence over other scales of policy-making, both processes promote voluntary, informal cooperation, potentially creating momentum for more integrated management of spatial development. Moreover, cooperation in the two large-scale regions derives from similar ideas about the need for seeing the 'bigger picture' in addressing challenges related to society and the environment. Both regions also entail a transboundary element, although the borders within So-Cal and EUSDR show varying degrees of permeability, as the analysis in the following sections highlights. Lastly, the two regions were selected because of their different treatment in the academic literature with a view to soft planning. While macro-regions in Europe have been conceptualized as soft spaces before, the concept has not been applied to understand spatial governance in the North American context.
The analytical framework builds primarily on two key sources for understanding planning and policy-making processes. In the first of these sources, Healey (2007) proposes that planning arenas and their underpinning structures can be analysed by paying attention to four aspects: (1) the networks and influence of participating actors, (2) the extent to which different stakeholders are involved, (3) the policy agendas and discourses which frame debates, conflicts, interests and strategies, and (4) the routines and repertoires for acting which structure day-to-day interactions and which shape how discourses and practices are changed and diffused. In the second source, Knoepfel et al. (2011) distinguish between four dimensions of public policy analysis: actors (political-administrative authorities, target groups, intended beneficiaries); rules (general and policy-specific); resources (e.g. personnel, money, political support, time, infrastructure); and policy content (substantive and institutional).
Building on the above categorisations, our analysis of the two cases addresses three dimensions: (i) agendas and discourses; (ii) actors; (iii) rules, routines and resources. In effect, these three dimensions relate to basic questions about why, who and how. Additionally, the article addresses the spatial delineations of the two regions. We are well aware that our research design cannot adequately describe all aspects related to So-Cal and the EUSDR in their entirety, as these large-scale cooperation processes are in themselves highly complex and multi-faceted. Nonetheless, we regard a comparison along these dimensions useful in order to understand under which circumstances soft planning processes unfold in the two regions.
As regards methodology, the article draws on the analysis of academic literature and policy documents relating to megaregions and macro-regions more generally and So-Cal and EUSDR more specifically. In the So-Cal case, the analysis is based on (1) strategy and policy documents with a megaregional perspective published by the Metropolitan Planning Organisations, the Regional Plan Association and the Department of Transportation as well as (2) a small number of stakeholder interviews conducted in spring 2018. For the EUSDR, the policy documents analysed include (1) the strategy documents of the Danube Region from the European Commission and Council, (2) the minutes and presentations of meetings in thematic sub-groups, and the (3) reflection papers and evaluations of the Danube Strategy Point, altogether accounting to more than 70 documents. Moreover, in the context of the EUSDR, one author has been involved in two large empirical studies including stakeholder interviews and workshops (Larrea et al. 2020;Sielker 2017), as well as in two wider policy processes in which the author assumed an advisory role.

Planning context in the United States
Responsibility for land-use planning and zoning in the US lies with the federal states (Knaap, Nedović-Budić, and Carbonell 2015). However, states delegate a majority of tasks to the local governments, as planning issues are mostly viewed as local concerns. After a failed attempt to establish a National Land Use Policy in 1970, the federal level continues to have no direct control over land-use issues. However, since the 1990s the federal government has enacted several laws that influence local land-use decisions, indicating that 'the federal government believes that it has some role to play in the land use regulatory process' (Salkin 2015, 33). Moreover, several federal agencies influence local land-use either through incentives or regulations (e.g. Department of Defense, Department of Transportation, Department of Housing and Urban Development).
There is considerable variation between US states regarding the roles of state and local administration in planning (Seltzer and Carbonell 2011). Generally, the influence of the state level has increased during the last decades and more attention has been paid to achieving regional coordination and consistency of neighbouring local plans (Salkin 2015). Since the 1990s, the notions of smart growth and sustainable development have been drawing attention to the inefficiencies of primarily local land-use policies. As a result, several states promote strategic planning and increased coordination of policies at the regional and state level while retaining implementation at the local level (Teitz and Barbour 2007). Currently, regional planning is mainly carried out by Metropolitan Planning Organizations (MPOs) at the metropolitan or city-regional scale. MPOs are expected to address issues such as climate change or sustainable transport and ensure vertical and horizontal integration of different policy levels and fields without having any legally established land use controls (Knaap, Nedović-Budić, and Carbonell 2015).

US megaregions as framework for growth management
The initial debate around large-scale urban agglomerations in the US started with the idea of the 'megalopolis' in the 1960s (Gottmann 1961). Since then, the concept of megaregions surfaced on several occasions in the US context (see Fleming 2015, 210). After a resurgence of interest in regional planning in North America in the 1990s (Wheeler 2002), megaregions have been increasingly discussed in the context of spatial planning and planning research (Dewar and Epstein 2007;Ross 2009). This interest was partly triggered by the US Bureau of the Census forecasting a 40% population increase by 2050, with a major share of growth in the biggest agglomerations (Dewar and Epstein 2007). In this context, the national infrastructure planning and policy programme 'America 2050' furthered policy debates on megaregions and identified eleven megaregions as illustrated in Figure 1 (Regional Plan Association 2006). The overarching goals for megaregions were to foster prosperity, equity and sustainability, in addition to promoting 'a new financing and decisionmaking framework that incorporates a variety of organizations and funding mechanisms' (Carbonell et al. 2005, 33).
America 2050 defines megaregions as 'large inter-connected or "networked" metropolitan areas' (Carbonell et al. 2005, 19). Dewar and Epstein (2007) describe them as characterized by 'environmental systems and topography, infrastructure systems, economic linkages, settlement patterns and land use, and shared culture and history' (113). America 2050 presented several potential governance models for megaregions, ranging from community-based bottom-up approaches to statemandated governance structures. The establishment and policy priorities of individual megaregions, however, was ultimately the task of actors in each megaregion. Consequently, the activities in different megaregions vary greatly. Recently, the Department of Transportation financed a research project on the potential of megaregions as new scales for transportation planning in practice (Federal Highway Administration 2013; Read et al. 2017).

The Southern California megaregion (So-Cal)
Spatial delineation: The spatial delineation of the So-Cal megaregion is fuzzy, as one interviewee from the San Diego Association of Governments (SANDAG) points out: There are many definitions for the megaregion and many geographical boundaries. […] There is no set definition, it continues to evolve. Sometimes academics make the definitions, sometimes others.
All delineations include the metropoles of Los Angeles and San Diego, as well as the surrounding and neighbouring counties (Imperial, Kern, Orange, Riverside, San Bernardino and Ventura). Some definitions consider the city of Tijuana or the region of Baja California in Mexico to be part of So-Cal, while others expand the megaregion to the East to include Las Vegas (Nevada) or Phoenix (Arizona). Depending on the delineation, So-Cal is home to between 20 and 30 million inhabitants, making it the third most populous megaregion in the US (Hagler 2009) and the world's tenth largest economy (Ross 2008).
The territories framing specific actions associated with the megaregion are mostly defined in an ad hoc manner matching the challenges at hand. In practice, cross-border cooperation remains an ambiguous aspect in So-Cal. Although there is little doubt about the functional connections between the US and Mexico (including education, job markets, trade and environmental issues), the city of Tijuana and the region of Baja California are seldom included in implementation strategies. Additionally, borders within the US between counties and cities continue to hinder cooperation. One attempt to reduce the obstacles associated with borders in So-Cal is the establishment of the Borders Committee through SANDAG (see below).
Agendas and discourses: The planning challenges associated with the continued rapid urbanization and population growth in California represent the dominant narrative behind So-Cal (Ross 2008). Moreover, So-Cal is considered important for attaining statewide climate goals that require coordination across municipal and regional boundaries, as one interviewee from the Local Agency Formation Commissions (LAFCO) highlights: These climate change plans could be the beginning of regional planning and megaregional planning in California. […] We will see what happens, it might still take 15 or even 30 years.
[…] The thresholds cannot be achieved unless there is megaregional coordination, so the state's effort towards climate change is one of the biggest drivers for cooperation.
The mega-regional scale is also relevant for the provision and coordination of public services, specifically transport and water infrastructure. Water management in California involves a multitude of different authorities (Hughes and Pincetl 2014), who cooperate at the megaregional scale in the Metropolitan Water District of Southern California, serving more than 19 million people (Erie and MacKenzie 2010), although the district was established before planners showed an interest in the concept of the megaregion.
So-Cal is also coined by multi-faceted and occasionally contradictory discourses regarding transport. On the one hand, the traffic volumes create problems such as congestion and air pollution. On the other hand, truck traffic, goods movement and logistics, especially to and from the region's main ports, are among the most important economic pillars of the region (Ross 2008). Debates at the mega-regional scale have also addressed high-speed rail connections. The recent rail investments (Pope 2018), supported primarily by the state of California, could bring about significant improvements regarding connectivity, travel times and carbon emissions in So-Cal. However, as the first phase of construction aims to connect San Francisco and Los Angeles, the improved rail connection could also shift the focus away from So-Cal with its functional relations to Mexico, and instead emphasize development in Central California.
Actors: Cooperation in So-Cal is not based on a defined governance structure but relies on the voluntary participation of public and private actors. This results in the fragmented involvement of interested actors, cooperating in different forms related to specific issues. Currently, MPOs are the main drivers of mega-regional cooperation, although they have no legal mandate for mega-regional planning, and differences in commitment to and interest in So-Cal can be observed between the MPOs in the region. In 2005, the Southern California Association of Governments (SCAG), SAN-DAG and the Kern County Council of Governments began collaborating on a planned growth strategy for So-Cal (Kern County Council of Governments et al. 2005;Ross 2008). SANDAG included the mega-regional perspective also in their other plans, for instance in the regional plan San Diego Forward (SANDAG 2015). In addition to highlighting the economic interdependencies and opportunities for interregional transport, the plan emphasizes 'the role that regional planning agencies can play as conveners, providing a forum to discuss interregional and megaregional planning issues and facilitating interregional planning efforts' (Read et al. 2017, 17).
Apart from the US-wide efforts towards mega-regional cooperation highlighted earlier (Carbonell et al. 2005; Federal Highway Administration 2013), there is little interest in So-Cal from the federal or state level. Moreover, the perceived importance of mega-regional planning is relatively low among cities and local governments. Regional planning organizations (mostly MPOs) are more inclined towards cooperation at the mega-regional scale but their position and political weight vis-à-vis local authorities remains weak.
Rules, routines and resources: Planning and politics in California are characterized by an individualistic paradigm and an emphasis on decision-making at the local level (Jonas and Pincetl 2006). Planning at the regional and state level is generally met with scepticism or even mistrust (Teitz and Barbour 2007). Planning at the mega-regional scale is institutionally even weaker than state planning, as there are no agreed upon rules or practices regarding how mega-regional cooperation relates to activities and policies at the local or regional level, as one interviewee from SANDAG claims: Usually it is so that the federal government or state government gives a framework and then the policies are developed and implemented at the local or regional level. But with the megaregion, there is no framework. So, the situation is so that those who want to participate, participatebut not everyone does.
[…] Instead, those that see value pick it up and run with it.
[…] That's why megaregions are in many ways still more of a concept.
Actions following debates related to climate change, service provision and transport in So-Cal are thus entirely dependent on the voluntary cooperation of different actors in individual projects. While this provides the freedom of finding ways of cooperating on a case-to-case basis, it also means that negotiation needs time and realization of projects remains uncertain.
One example for establishing mega-regional routines is the Borders Committee, initiated by SANDAG (Read et al. 2017). The committee aims to facilitate cooperation across the international border between Mexico and the US, the administrative borders between cities and counties within California, as well as the cultural and institutional borders faced by Native American tribal governments. Despite its informal character, the committee has been successful in implementing cross-border projects, for example, the development of a strategic plan for the Otay Mesa area between San Diego and Tijuana (SANDAG 2007(SANDAG , 2012. Although SANDAG has no mandate to extend its planning area across the border, a joint plan was developed together with Mexican authorities which addresses issues related to transportation, housing, economic development and environment. There are no immediate financial or personnel resources associated with So-Cal. Public-private partnerships are regarded as potential financing model (e.g. for transport infrastructure), highlighting the potential of So-Cal to bridge different interests (Kern County Council of Governments et al. 2005). However, these attempts are not coordinated within comprehensive funding schemes.

Planning context in Europe
Planning across the EU differs more widely than across the US (Faludi 2015), although the overall division of planning competences is broadly comparable, as the European Union level holds no competence for land-use planning; this is a sovereign task of member states. Nonetheless, the EU potentially affects land-use planning and territorial development through at least three types of policies (Purkarthofer 2016). First, the EU uses financial instruments and offers co-financed funding schemes, and especially funds associated with EU Cohesion Policy are potentially influential for spatial development. Second, the EU shares the responsibility for several sectoral policy areas with its member states, such as environment, transport, energy and agriculture (Sielker 2018), for which the EU can enact regulations with legal implications. Third, the EU or representatives from member states can endorse strategies which have implications for spatial development, for instance strategies for energy supply networks and transport infrastructure networks. These discursive influences can also entail the transfer of non-binding policy guidelines, benchmarks or 'best practices' via the EU level.
While governance arrangements and planning systems differ between nation states, all states have some form of national, regional and local administration with responsibilities for planning (Nadin et al. 2018). The national level, traditionally more important in centralized states, typically enacts planning legislation, provides strategic guidelines and plans, and discusses and negotiates planning at the EU level and with other countries. The regional level, traditionally more independent in federal states, is required for the implementation of EU regional and cohesion policy and in some cases, regional land-use planning. The local level is typically responsible for land-use planning and zoning. 5.2. EU macro-regional strategies as collaborative approaches for territorial cooperation EU macro-regions are cooperation frameworks through which EU member states and non-member states address common challenges by developing joint strategies and outlining priorities for joint activities (European Commission 2013). To date, the EU has endorsed four macro-regional strategies: the Baltic Sea Region (in 2009), the Danube Region (in 2011), the Adriatic and Ionian Region (in 2015) and the Alpine Region (in 2015). Macro-regions cover vast geographic areas defined by large-scale topographical features, specifically sea and river basins and mountain ranges as shown in Figure 2 (European Commission 2020). Each of the macro-regional strategies is characterized by 'three no's': no new legislation, no new administrative structures and no new funds (Sielker 2016a). Instead, macro-regions are tasked with ensuring better use and coordination of existing institutions, organizations and funding opportunities.
Despite these limitations, macro-regions represent new soft spaces on an intermediate scale between the nation state and EU level. Macro-regional cooperation builds on a transnational governance structure, in which governmental actors from different countries come to consensus-based agreements while the European Commission takes up a mediating and coordinating role. The countries develop a joint strategy and define thematic areas on which they intend to cooperate. In order to achieve the goals defined in the strategy, macro-regions rely mainly on project-driven implementation supported by European or national funds.

The EU macro-regional strategy for the Danube region (EUSDR)
Spatial delineation: The EUSDR covers ten countries along the Danube River and additionally includes four countries within the river's water catchment area. The EUSDR comprises nine EU member states (Austria, Bulgaria, Croatia, Czech Republic, parts of Germany, Hungary, Romania, Slovakia, and Slovenia), two EU candidate countries (Montenegro and Serbia) and three countries that do not have immediate prospects of joining the EU (Bosnia-Hercegovina, Moldova and parts of Ukraine). Altogether, this constitutes a region of approximately 115 million inhabitants and an area of around 800,000 km 2 . This broad territorial coverage does not apply to all interventions. Instead, the spatial delineations vary for individual themes and actions (Sielker 2017), as noted by a Priority Coordinator: The territorial coverage of projects greatly depends on the topics addressed. While some working groups focus in their work on the riparian states of the Danube river, others focus on specific parts of the Danube Region, e.g. the challenges of the Danube Delta. Also, the focus of implementation activities changes over time. [authors' own translation] The territorial coverage of the EUSDR follows the idea of a shared 'functional space' defined by the river basin and the resulting need for coordination, definition of joint political priorities and targeted steering of funds. The novelty of the territorial coverage and its fuzziness becomes also apparent considering the coordination between EU and non-EU countries, which have different access to (EU) funding opportunities (European Commission 2020).
Agendas and discourses: The idea for an EU Strategy for the Danube Region first emerged in 2008 primarily due to environmental and navigation problems on the Danube River. Following a 2year drafting process the EUSDR was formally adopted in 2011 and cooperation themes were agreed upon. Cooperation focuses on four thematic pillars ('Connecting the Region', 'Protecting the Environment', 'Strengthening the Region' and 'Building Prosperity') and twelve priority areas, most of which can be considered to some degree relevant for planning. In May 2020, an updated EUSDR Action Plan was published, foreseeing the continuation of the twelve Priority Areas, indicating a lock-in of the broad topics addressed.
Actors: The EUSDR builds on the principles of multi-level governance, which are also visible in its actor constellation. One national coordinator per country acts as key person responsible for thematic steering of the strategy and for ensuring links between the EU, the national and sub-national levels. For each priority area, an international steering group, consisting of representatives from national ministries, is responsible for devising implementation activities. The steering group develops targets for the priority area, establishes a work plan and selects projects and implementation activities. Concrete actions and projects are then realized with the help of working groups, which involve actors from the public and private sector, such as representatives of cities, companies or NGOs, on a voluntary basis (Sielker 2016a;Wulf 2015). In the EUSDR, national stakeholders are key actors shaping agendas and establishing links with funding programmes. The European Commission supports the macro-region with coordination, mediation and communication services (Plangger 2018).
Despite the intention of creating a level playing field between member states, conflict-laden power relationships between states prevail, as a representative from the European Commission notes: Intergovernmentalism remains the principle for cooperation. Yet, cooperation in a macro-region also means to overcome existing power relations between richer and poorer countries, as well as between EU Member States and Non-EU countries. The EUSDR helps overcome some existing boundaries. Yet, sometimes it reinforces the perceived challenges.
Rules, routines and resources: The governance structures associated with the EUSDR can be characterized as nested with no strict hierarchies in place. The EUSDR relies on informal and voluntary cooperation and consensus finding between different levels of government. While private companies and civil society actors have the possibility to be involved in the implementation of activities, the macro-region builds on existing governmental structures for its overall functioning. National actors have considerable leeway in deciding who will take up certain roles related to the EUSDR in their country.
The EUSDR can be understood as an arena for strategic framing of debates and international coordination at the macro-regional scale, while implementation is foreseen at multiple spatial scales and through various mechanisms. Planning specifically is perceived as a set of interconnected processes at different scales, which are not steered but supported through the EUSDR. The most obvious connections between the EUSDR and planning lie in the transport, environment and energy sectors. A concrete example is the contribution of the EUSDR to changes in eligibility for funding in the Connecting Europe Facility, which enabled the realization of dredging activities in the riverbed that significantly improved shipping conditions and navigability of the Danube River.
The macro-region is devised not to create new legislation, institutions or funding. However, synergies and alignment with the EU multi-annual funding frameworks are crucial to ensure the practical operation of the EUSDR. The transnational INTERREG programme for the Danube Region is used to finance the Danube Strategy Point (hosted by the City of Vienna and Romania), which takes up a major role regarding coordination in the macro-region. The role of macroregional strategies as a guidance framework for the use of EU funding is strengthened in the proposals for the 2021-2027 funding period.

Comparing So-Cal and EUSDR: the influence of soft planning
Spatial delineations: Both case study regions can be characterized as soft spaces, representing new and flexible spatial delineations, yet they follow different logics. In So-Cal, various different megaregional spaces exist, invoked by different actors and used to frame different cooperation process and themes. In the EUSDR, the 'outer borders' of the macro-region are clearly defined (though not impermeable), but specific projects and actions use varying 'sub-spaces' within these borders.
Agendas and discourses: Both So-Cal and EUSDR represent large-scale spatial delineations and make use of appealing spatial imaginaries, which can play an important role in bringing stakeholders together that identify with these soft spaces by giving it a name and an image (Davoudi et al. 2018;Walsh 2014). However, the narratives used to construct these imaginaries differ substantially. The idea of So-Cal is based on the necessity to accommodate the expected population growth and to solve problems related to infrastructure provision and environmental protection. The transboundary element is present in the San Diego-Tijuana sub-region but initiatives from state or federal level often omit the involvement of Mexico in mega-regional cooperation. While cities in Nevada and Arizona are sometimes considered as part of So-Cal, examples of inter-state cooperation under the umbrella of the megaregion are sparse. In the EUSDR, the joint use of the Danube river basin and catchment area as topographically unique space served as the initial narrative fuelling the establishment of the macro-region. Transboundary cooperation is a structuring principle, as geopolitical concerns and environmental issues are among the drivers for collaboration in the region. The choice of cooperation themes is more systematic in the EUSDR, resulting from a long-term negotiation process, in comparison with So-Cal, where individual cooperation themes are identified in an ad-hoc manner.
Actors: The importance of transboundary cooperation is also reflected in the governance approach of the EUSDR, in which steering groups and working groups are composed of representatives from different countries. In addition, national coordinators from the participating countries ensure commitment and add political weight to the macro-region despite its voluntary character. The governance approach employed in So-Cal is less institutionalized and does not include nested structures comparable to the EUSDR. The reliance on regional authorities, especially MPOs, to further mega-regional cooperation reflects the low political importance ascribed to the megaregion. Thus, while both examples are characterized by scattered responsibilities for planning-related issues, the EUSDR provides structures for cooperation across sectoral and governmental levels. Despite representing a soft space with previously 'thin' transnational institutions (Dühr 2018), the EUSDR is therefore institutionalized to a considerably higher degree than So-Cal, due to the systematic involvement of a multitude of actors.
Rules, routines and resources: Both So-Cal and EUSDR rely on voluntary and legally non-binding cooperation, however, from an institutional-procedural perspective, they show considerable differences. Cooperation in the EUSDR is based on a set of informal yet clearly stipulated rules and thus resembles a coherent governance structure, while generating new actor constellations and power relations. An official endorsement and continuous administrative support by the EU adds to the degree of institutionalization of the EUSDR. The EUSDR is primarily used as an arena for debate, mediation and strategic coordination between various actors and levels of government in different countries. This resonates with the understanding of planning as interconnected processes at different scales in the EUSDR.
Cooperation in So-Cal is driven by individual themes or projects, and the mega-regional scale is not systematically linked with existing structures and processes. So-Cal is thus an appealing spatial imaginary framing this soft cooperation, especially to interest groups such as the Regional Plan Association or regional authorities who face challenges posed by hard borders. However, weak governance arrangements, fragmented responsibilities for different aspects related to infrastructure provision and strong local control regarding planning make mega-regional cooperation especially challenging. In both regions, scarce financial and personnel resources present a hindrance to extensive cooperation. Within the EUSDR, actors have the opportunity to jointly apply for funding, for example via EU Cohesion Policy, which creates an additional incentive for cooperation. In So-Cal, no comparable framework of funding opportunities exists.

Lessons from soft planning in So-Cal and EUSDR
By comparing the two examples of large-scale cooperation presented in this article, it has been possible to shed light on the question of how soft spaces and soft planning not only represent an additional scale in the realm of planning but also contribute to policy development and delivery. We argue that these new large-scale cooperation areas can positively affect spatial development and identify four lessons to be drawn from comparing So-Cal and EUSDR: (1) Soft planning depends on perceived joint gains Establishing processes of cooperation within new spatial delineations is not an end in itself. Such processes need to address an exigency and give the promise of a joint gain that is desirable for all actors involved. Transboundary cooperation in the case of the EUSDR clearly is such an incentive, which is especially appealing to non-EU states, as they do not have that many opportunities to participate in multilateral processes and discussions. The recent, and partly still ongoing, conflicts and political disagreements in the Danube region have created tensions between specific countries and regions, which are easier to put aside in a setting involving several countries. The opportunity to obtain EU funding for joint projects framed under the EUSDR is yet another, and rather tangible, joint gain. However, it should be kept in mind that the funding amounts vary, and the funds associated with the territorial cooperation programmes are limited in comparison with national funds and sectoral funding sources such as the Connecting Europe Facility. Nonetheless, the prospect of obtaining funding is an easily agreeable goal fuelling joint activities, which can in turn set the wheels in motion for deeper exchange and long-term cooperation.
(2) Soft planning is not the goal but can guide the way Whether perceived joint gains are in fact achieved might not be crucial for assessing the added value of soft planning. More important than the outcomes of soft planning processes are the processes themselves. Through deliberation and collaboration, they provide an opportunity to develop future agendas and frame problems. This can in turn provide a strategic push, necessary to enable or steer planning at other spatial scales. This aspect reveals undeniable parallels between soft planning and strategic spatial planning (Albrechts 2004), and relates to the discourse on performance in planning (Faludi 2000). In the case of large-scale cooperation as discussed in this article, it is almost impossible to consider implementation as the goal. Instead, the examples show that soft planning can bring together actors that previously did not have a joint arena for deliberation. The example of the EUSDR gives evidence as to the impacts cooperation has on regulations at other scales (Sielker 2016b(Sielker , 2017. (3) Soft planning relies on rules, even if they are not binding The analysis shows that even if soft planning is not regulated by law, meaningful cooperation can only be achieved when a joint understanding of 'how things are done' is established. Such rules need to be flexible or innovative enough to break away from existing patterns or stalemate situations. In the case of the EUSDR, the negotiation of these rules and working ways formed part of the negotiations preceding the establishment of the strategy. Although the European Commission accompanied and endorsed the development of the EUSDR, the participating countries were actively shaping the institutional architecture of the macro-region. A comparable process did not take place in the context of So-Cal, thus mega-regional activities remained single incidents rather than adding up to a broader strategy.
(4) Soft planning needs some anchorage in hard spaces In order to provide contribute to policy development and delivery, soft planning cannot be completely detached from the administrative system. This is especially apparent regarding the actors involved in soft planning. Actors with specific subject and content knowledge and those with decision-making powers are needed to ensure both the political support and practical application of large-scale soft planning. However, soft planning can contribute to broaden the scope of actors involved in such processes and bring together actors in new constellations not defined by government levels, and thus support the generation of new ideas. The working groups in the EUSDR are a good example of involving representatives from national, regional and local authorities, responsible for different sectoral policies, as well as NGOs and the private sector, working together in a collaborative and equal setting that otherwise would not exist.

Conclusions
This article has juxtaposed macro-regional and mega-regional cooperation in Europe and North America from the perspective of spatial planning and governance. The comparison of Southern California and the Danube Region has revealed differences between the two approaches but also highlighted comparable characteristics. Both outline new large-scale geographies and spatial imaginaries, which can be understood as examples of soft spaces (cf. Allmendinger and Haughton 2009). Cooperation in these soft spaces can be characterized as soft planning (cf. Faludi 2010), bringing about new forms of voluntary coordination, negotiation and learning in informal settings between different actors, while facing a fragmentation of competences and scarce personnel and financial resources.
The cooperation arrangements in So-Cal and the EUSDR show that soft planning is not about a search for the 'holy grail', aimed at finding the right scale to resolve challenges and providing solutions. Rather, it is about finding ways to think 'outside the box'. The box in this case is an equally suitable metaphor for the bounded container view on space, the sectoral silos of government and established ways of policy-making. The analysis suggests that soft planning can contribute to territorial integration when a perception exists that there are joint gains to be made, when informal rules are negotiated in context-specific and bottom-up processes, when soft spaces are used as arenas of deliberation to renegotiate shared agendas, and when actors succeed in ensuring the anchorage of informal cooperation in other arenas.
The analysis also concurs with the early work of Metzger and Schmitt (2012) who ascribe importance to the hardening of delineation and governance of soft spaces over time. Such a process of hardening can be observed in the EUSDR, where the network of stakeholders is becoming consolidated and it is now more difficult to involve new stakeholders than in the past. Moreover, the analysis reveals that the spatial imaginaries used to corroborate the narrative have contributed to a hardening of cooperation arrangements. The temporal dynamics suggest that the rules, routines and spatial delineations are more prone to harden after an initial 'experimental impulse' and a 'macro-regional hype' (Allmendinger, Chilla, andSielker 2014, 2711). However, agendas and discourses which are part of the process of soft planning remain open to change. While the spatial imaginary is often used to bind the stakeholders together in the initial phases of the development of a macro-regional strategy, it increasingly turns into a 'marketing tool' over the course of time.
In both regions, a few parties act as the main catalysts to advance cooperation, although the political weight of these actors differs considerably between the two cases. While the analysis did not permit to investigate the actors' intentions in detail, it revealed that actors use the mega/macroregional arena to further the interests of their organization or jurisdiction. Instead of 'megaregional planning', i.e. comprehensive plans at the mega/macro-regional scale (cf. Harrison and Gu 2021), diverse connections can be seen between the large-scale cooperation frameworks and specific interventions in various organizational settings at lower spatial scales.
While it is unsurprising that actors are not pursuing mega-/macro-regional interests as such, there is a danger of a few powerful voices dominating cooperation in large-scale soft spaces such as So-Cal and EUSDR due to uneven power relations. This reinforces the importance of understanding the lessons from soft planning that are identified earlier in this article. While soft planning can serve as an arena for deliberation and negotiation in order to bring together various actors, decision-making competences should remain with governance actors in 'hard spaces' unless mechanisms for ensuring democratic legitimacy are found, which would undoubtedly lead to a hardening of these processes. Nevertheless, even without formal competences, the discursive appeal of the spatial imaginaries associated with large-scale soft spaces can be considerable.
The comparison of the two cases reveals that the EUSDR has led to deeper cooperation, although we do not want to claim that all aspects of cooperation work equally well and that the process is without difficulties. Evaluating the success of the EUSDR goes beyond the scope of this article and is a challenging task, mainly because direct causal relations cannot easily be detected in such a complex setting. However, recent research has shown that the activities and agendas set through macro-regional debates have influenced decision-making at other scales despite the fact that macroregional cooperation is not associated with formal decision-making competences (Gänzle et al. 2018;Plangger 2018;Sielker 2016a).
According to the analysis presented in this paper, So-Cal has not yet fully revealed its added value and thus remains a largely toothless spatial imaginary. Certainly, So-Cal has not succeeded in serving as a new financing and decision-making framework bringing together various organizations and funding mechanisms (cf. Carbonell et al. 2005, 33). One explanation could be that the real need for cooperation and potential joint gains at this scale have not yet become apparent. However, this may change as the region continues to face high population growth, increased commuter traffic and goods transport, and growing environmental burdens, especially around the border between Mexico and the US. While So-Cal might not have the political weight to resolve these problems, the megaregion could be a suitable arena to discuss strategic concerns and create the momentum needed to persuade other levels of government to take strategic action. The Borders Committee, which was established in the San Diego and Tijuana area, is an example of the creation of an arena for deliberation and collaboration, albeit at a significantly smaller scale.
To be clear, this paper does not wish to advocate the direct transfer of the EU macro-regional approach to other large-scale regions around the globe. Instead, it intends to highlight that soft planning is only meaningful when the processes and rules it follows are negotiated in a contextspecific and bottom-up process. Top-down guidance might be helpful and needed to support and frame such a process. Agendas, actors, rules and resources need to be present and somewhat aligned in search for joint solutions to transboundary challenges. If this is the case, soft planning can make a contribution to transforming megaregions from an academic concept into a framework for meaningful cooperation, which can ultimately improve the coordination of spatial development at other scales.

Disclosure statement
No potential conflict of interest was reported by the author(s).