The Legitimation of Planning Processes as a Challenge to Metropolitan Governance

: This study identiﬁes three types of legitimation from the literature that can be applied within metropolitan governance in the contested sphere of spatial planning: input legitimation, throughput legitimation, and output legitimation. The reason for discussing di ﬀ erent forms of legitimation within metropolitan governance is that, globally, only a relatively few metropolitan regions are governed directly through a single elected tier of government such as a regional council. Thus, governance mechanisms in most metropolitan regions involve some form of joint working or cross border governance initiatives that have to be legitimized in the absence of a single overarching elected council covering the whole metropolitan area. The main question discussed in this paper is, therefore, whether all three types of legitimation identiﬁed are utilized to legitimize governance mechanisms at the metropolitan scale with a speciﬁc focus—as a core part of metropolitan governance—on spatial planning processes and projects. In conceptual terms, our typology structures fuzzy lines of legitimation across the three (the “how”, “who” and “what”) suggested aspects of metropolitan governance in the literature. From this point, we draw on cross-case reviews of variables involved in the design, application, and outcome of input, throughput, and output legitimation in Germany and England, chosen because neither has a formal tier of metropolitan-wide government despite their di ﬀ erences in terms of their highly regionalised and highly centralised national government contexts respectively. This relational methodology helps us to learn about the contextual dynamics of how the three types of legitimation might reinforce one another in di ﬀ erent international settings, leading to the overall conclusion that they will work best in combination, although output legitimation has a distinctive capacity to work in less formal settings.


Introduction
Currently, more than half of humankind, around 3.5 billion people, live in cities. According to predictions of the United Nations, this figure is likely to surpass 6.5 billion people by 2050. This creates challenges with regard to how urban spaces can and should be designed and how they are to be governed (United Nations 2014). Employment possibilities and the availability of functioning infrastructure are factors favouring migration into cities (Ziafati Bafarasat 2017). However, this concentration of people and possibilities in cities needs to be organized (Zimmermann and Heinelt 2012). Although common to urban areas more generally, these challenges are often particularly acute 1.
What is the primary objective of the governance mechanisms: to create an external profile or internal control? 2.
To what scale do governance mechanisms refer: diffuse or clear and to a wide or narrow geographical delimitation of the metropolitan region? 3.
What is the scope of the governance mechanisms: a functionally specialized or a territorial-multifunctional orientation? 4.
Do the governance mechanisms follow a hard or soft institutionalization? How are legal constitution, competencies, resources, and political legitimation organized? 5.
Are governance mechanisms institutionalized or are governance mechanisms played out through the consultation and mobilization of private and civil society actors? 6.
What instruments are used in governance mechanisms: formal and regulative instruments or communication-based instruments aiming to convince relevant actors? 7.
What period do governance mechanisms cover? Do they govern through projects or through an integrative control approach to spatial policy? 8.
How is metropolitan governance organized as multi-level governance: are local and regional levels tightly or loosely coupled and is external control carried out through norms or symbols?
All of these aspects are important in metropolitan governance and can differ in each region, consequently leading to a broad variety of potential forms of metropolitan governance. These complex anatomies call for corresponding policy informing investigations to enable more effective governance of the emerging "metropolitan revolution" (Katz and Bradley 2013). Meanwhile, the UN-Habitat (2012) highlights spatial planning as the crux to achieving sustainable prosperity; its integrative perspective can overcome sectoral policy silos by recognising and balancing the nexus between different dimensions of urban functionalities that are most densely in play at the metropolitan level (Wong 2015). Spatial planning aims to coordinate the spatial impacts of sectoral policies and is an important lever for promoting sustainable development and improving the quality of life (United Nations 2008)-core agendas for public policy that are otherwise very difficult to address in the frequent absence of a single metropolitan or regional government (Ziafati Bafarasat and Baker 2016a). It has been undertaken to serve different purposes including economic development and restructuring, place branding, core city revival, flood risk control, institutional and governance reform, mega infrastructure development, and trans-territorial policy coordination (Albrechts 2017), with international observations of some success at the metropolitan and regional levels (Healey et al. 2003;Albrechts et al. 2017). Spatial planning is therefore central to virtually all of the above aspects of metropolitan governance (Ziafati Bafarasat 2015) and, in particular, provides an integrative context to examine the "how", "who" and "what" of the fourth, fifth, and sixth aspects (Pugalis and Ankowska forthcoming). The legitimation of metropolitan spatial planning is thus a core topic at the intersection of these aspects that merits conceptual and empirical attention in the context of reduced trust in, and discontent with, neoliberal urban governance (Tait 2012;Kwok et al. 2018), issues of participation and accountability in the fuzzy spaces of metropolitan decision making (da Cruz et al. 2019), and the failures of soft spatial management to deliver on its promises of the common good .
In political science, legitimation in the narrower sense refers to the justification of a state for its sovereign or non-sovereign actions. With regard to metropolitan governance, we ask how spatial planning processes-as one part of metropolitan governance (Salet et al. 2015;Zimmermann et al. 2020)-are legitimated. Three basic types of legitimation are applied in the governance discourse in the EU (Schmidt 2013;Zimmermann 2014): input legitimation, throughput legitimation, and output legitimation.

•
Input legitimation is based on the normative principle of approval by the ruled (governance by majority consent). It is the predominant category of legitimation in law. Input legitimacy refers to the quality of the participatory process leading to binding political decisions. Elected politicians deciding on plans and strategies developed by planning professionals are understood to provide input legitimation. A key aspect is deciding about the public interest (Alexander 2002). Input legitimation requires not only councils consisting of elected officials to ensure accountability and responsibility but also professional staff providing administrative and technical capacity to prepare binding political decisions (Growe and Jemming 2019). One important factor here is a stable source for financing the implementation of political decisions. The fiscal stability of the governance mechanisms may be enabled by financial support through the federal or the state level. Additionally, many regional councils opt to raise funds through fees and taxes themselves (Wolf and Bryan 2009).

•
Throughput legitimation is based on the participation of those controlled in the governance process (governance with the people). Approaches in this direction are the forms of direct democracy such as popular initiatives or referendums. Such participation always presupposes the possibility of access for the participants to information, and thus administrative transparency and freedom of information. In planning processes, a reference to throughput legitimation is made with collaborative exercises (Barton et al. 2015;Suškevičs 2019). Collaborative exercises may involve individual citizen consultations (Hamilton 2013;Wagner et al. 2019) or policy making and project partnerships between economic, governmental, and non-government actors (Grigsby 1996;Ziafati Bafarasat 2016). Whilst individual citizen consultations expand throughput legitimation due to the larger number of citizens engaged in making quick choices between pre-determined options, policy making and project partnerships deepen throughput legitimation, although they are sometimes reflective of an elitist approach to stakeholder identification (Ziafati Bafarasat and Baker 2016b).

•
Output legitimation is based on the functional principle of utility (governance for the people). Actors who generate useful services are not necessarily always democratically elected or belong to a recognized government but are accepted by the controlled due to the provision of services. In planning processes, output legitimation is often discussed in the context of developing sustainable spatial structures, as sustainability is seen as goal of overriding importance (Counsell and Haughton 2006). With a recent surge of interest in spatial reconfigurations that stimulate economic competitiveness and the sustainable development of cities and regions, economic prosperity and environmental conservation establish another aspect of output legitimation for spatial planning, in particular at the metropolitan level (Ziafati Pugalis 2018, 2019). Another context for discussing output legitimation in planning processes is unusual spatial configurations like cross-border metropolitan regions where elected regional councils are difficult to find and participation is difficult to realise (Fricke 2014;Harrison and Growe 2014b;Decoville and Durand 2016).
The question of how spatial planning processes-as one part of metropolitan governance-are legitimated is discussed in this paper using the typology of input, throughput, and output legitimation. Especially in metropolitan regions without elected regional politicians, the question of legitimation is crucial (Growe and Jemming 2019). This observation is the background for developing the main question of the paper. Globally, only a few metropolitan regions are governed through an elected metropolitan or regional council and, therefore, governance mechanisms in most metropolitan regions involve joint working or cross-boundary initiatives which have to be legitimized without a single, overarching, elected regional council at the metropolitan scale. The main question discussed in this paper is, therefore, whether all three types of legitimation are utilized to legitimize governance mechanisms, with a specific focus-as a part of metropolitan governance-on spatial planning processes and projects. Due to the diverse range of governance mechanisms, it is anticipated that all three forms of legitimation identified in the literature will be found to be relevant in different cases, supporting metropolitan spatial planning processes as well as associated planning projects such as major metropolitan infrastructure.
The paper draws on the empirical literature and extensive hands on experience of the authors to compare regions from Germany and England using this threefold typology. It is acknowledged that there is no universal definition of what constitutes a metropolitan area. Therefore, starting from the theoretical-based differentiation of the three forms of legitimation, a systematic identification of potential metropolitan cases has been carried out for German and English regions (see Table 1). For example, in Germany a total number of 15 regions are identified as metropolitan regions in national spatial strategies. Out of these regions and based on empirical literature as well as our own empirical work, in terms of interviews and document analyses three cases have been identified. The three cases have been described in detail so that each of the cases can be understood from the outside. In England, the starting point was in Greater London and the (former) metropolitan county council areas of the midlands and northern England, as well as other predominantly urbanized Combined Authority (CA) areas. Within these, particular focus has been given to examples from Greater Manchester. The focus on both German and English regions allows the consideration of two countries with different traditions of strong regional government (Germany) against a more centralised approach (England) and how this might impact on the metropolitan scale. It also brings together the strongly theoretical-based English perspective and the (prevalent) applied research tradition of German Human Geography (Jöns and Freytag 2015). Jöns and Freytag (2015, p. 15) argue that "the engagement in German-language geographical research and teaching with visions and strategies of public planning [ . . . ] as well as with practitioners working more often in government-funded institutions of spatial analysis and planning than in private planning consultancies might be of particular interest to Anglophone audiences".
The rest of this paper is structured as follows: Section 2 analyzes legitimation as a challenge of spatial planning in German metropolitan governance. Section 3 analyzes legitimation as a challenge of spatial planning in English metropolitan governance. Section 4 systematically summarizes the results and Section 5 reflects on the challenge of legitimation as part of metropolitan governance.

German Planning Background
While Germany is known as a federally organized nation state where spatial planning is anchored strongly at the regional level, metropolitan governance in a broader sense gained importance during the 1990s (Blotevogel and Schmitt 2006;Blotevogel et al. 2014). Realizing that metropolitan regions have particularly important socio-economic functions and are seriously challenged through environmental and socio-economic processes as well, there has been an increasing awareness that the government structures and processes in these regions have to be renewed in order to meet the changed requirements (Blatter 2005). Thus, during the 1990s a variety of governance forms developed in Germany and, although objectives, scales, and forms differ, all governance mechanisms deal with the questions of spatial regional development and, thus, with questions of regional planning (Zimmermann 2014).
Although the national level has legislative capacities with regard to spatial planning, only a few sectors exhibit extensive planning competency on this level-e.g., traffic planning. When it comes to comprehensive planning, the development and implementation of plans for land use and infrastructure across a number of sectors takes place at the regional level (Blotevogel et al. 2014). Regional planning is integrated in a multilevel system of spatial planning. Plans outlining spatial structures and areas allowing or forbidding different land uses are developed as important instruments on all levels. Elected politicians decide on these formal instruments and professional planners prepare the plans in planning administrations on all three levels. These plans are named "spatial structure plans" at the state level, "regional plans" at the regional level, and "preparatory land-use plans" at the municipal level. Due to the mutual feedback principle within the multilevel system of spatial planning, a vertical integration of planning aims is obligatory. This principle outlines the countervailing influence between local, regional, and supra-regional planning (Pahl-Weber and Henckel 2008). For example, regional plans are prepared based on state spatial planning requirements. The spatial planning aims set forth in state spatial structure plans must be complied with and detailed in the plans drawn up for regions of the state. Meanwhile, when preparing regional plans there are also processes of horizontal integration, and integrating the aims of sectoral planning in cross-sectional, comprehensive planning, is necessary (Pahl-Weber and Henckel 2008).
Given that horizontal and vertical integration processes are obligatory when developing regional plans, the main concerns in regional planning in Germany deal with the effectiveness of regional plans. To support a regions' development, not only the determination of future land uses is necessary but also the implementation of these uses-e.g., through private investors or public financing. Also important in this context is that private actors often follow a different logic for land use and development than political actors (network logic versus territorial logic) (Harrison and Growe 2014b). Thus, a broader governance perspective becomes necessary, especially in closely interrelated and densely urbanized metropolitan regions (Schmitt and Danielzyk 2018).

Metropolitan Regions and Metropolitan Governance in Germany
During the 1990s, the focus on metropolitan regions was first evident in the policy documents produced by the Standing Conference of Ministries Responsible forSpatial Planning (Ministerkonferenz für Raumordnung). To develop spatial strategies for a reunified German territory, "Guidelines for Spatial Planning" (Raumordnungspolitischer Orientierungsrahmen) have been adopted, identifying the strategic importance of large metropolitan clusters in an extension of the existing system of central places (Harrison and Growe 2014a).
The idea of metropolitan regions as a potential category in spatial planning resulted from interaction between practitioners and scientists. Metropolitan regions, as the spatial basis for governance mechanisms, do not always refer to the existing delimitations of regional planning administrations. The Standing Conference of Ministries Responsible for Spatial Planning identified six agglomerations (Berlin/Brandenburg, Hamburg, Munich, Rhine-Main, Rhine-Ruhr, and Stuttgart) of "superior" strategic importance, with a seventh (Halle/Leipzig-Sachsendreieck) identified and subsequently added to the list two years later. These regions were to be known as "European Metropolitan Regions". However, a distinct spatial delimitation was no precondition to be nominated. The nomination referred to the main city (or main axis), allowing the regional actors to fill the label with their conceptions. Although the identification as metropolitan region was, and is still today, merely a label, many actors did believe that European Metropolitan Region status might lead to more direct subsidies (Blotevogel and Schmitt 2006). Thus, further four agglomerations (Bremen-Oldenburg, Hanover-Braunschweig-Göttingen-Wolfsburg, Nuremberg, and Rhine-Neckar) successfully aspired in 2005 to be admitted to this new politically constructed spatial category (Harrison and Growe 2014a). In 2016, four cross-border regions-The Lake Region (Europäischer Verflechtungsraum Bodensee), Euregio Maas-Rhein, Greater Region (Großregion SaarLorLux+), and Upper Rhine Region (Oberrheinregion)-were added as cross-border metropolitan regions and, moreover, expanded the idea of metropolitan regions in Germany (Harrison and Growe 2014b).
Against this background, the integration of metropolitan regions in the multi-level system of spatial planning in Germany differs. The delimitation of some metropolitan regions reflects administrative boundaries at the regional (Rhine-Neckar) or at the lander level (Berlin/Brandenburg), enabling them to form the administrative demarcation for formal regional plans. The delimitation of other metropolitan regions does not reflect administrative boundaries, resulting in the need for various actors to actively aim to develop metropolitan regions through governance mechanisms such as networks, partially beyond regional and national borders as well. Summing up, the concept of metropolitan regions promotes the complementary combination of land use-oriented formal planning instruments and development-oriented governance mechanisms (Knieling and Blatter 2009;Zimmermann and Heinelt 2012;Zimmermann 2014).

Input Legitimation in Spatial Planning in Germany
Input legitimation in spatial planning as part of metropolitan governance is explained using the example of the European Metropolitan Region Stuttgart (EMRS) and the Verband Region Stuttgart. Both organizations refer to Stuttgart as their urban core; however, they are not identical. Put briefly, the Verband Region Stuttgart covers a smaller regional demarcation than the EMRS. However, the Verband Region Stuttgart is a regional multi-purpose association in combination with a directly elected assembly, while the EMRS is a loose network consisting of five regional associations-one of them being the Verband Region Stuttgart.
The Verband Region Stuttgartincludes 179 cities and municipalities from the five districts of Böblingen, Esslingen, Ludwigsburg, Göppingen, and Rems-Murr as well as the state capital of Stuttgart. Together, they form a regional agglomeration that is closely intertwined with economic relationships, commuter movements, and settlement structures in one of the most prosperous parts of Germany (Wagner and Growe 2020). However, as the challenges of the agglomeration do not stop at municipal or district boundaries, the region has joined forces and-supported by the Minister-President of that time-founded the Verband Region Stuttgart in 1994 with its directly elected regional assembly (Kübler 2012).
The Verband Region Stuttgart is one of only two regions in Germany with an elected regional assembly. 1 The regional assembly is part of the Verband Region Stuttgart, being the political entity for the region. The regional population decides every five years who will be the delegates in the regional assembly. The reasons for introducing a directly elected assembly were the expectations that this assembly might support a stronger sense of regional identity and strengthen the support for decisions that might confront the included municipalities (Zimmermann 2014).
Parties and voter associations nominate candidates for the regional assembly. The regional assembly has 80 members; with compensation seats, it can host a maximum of 96 elected officials. Formerly dominated by local politicians (predominantly majors of the included municipalities), the members of the regional assembly today have different political backgrounds. Although local politicians from the included municipalities still can be found, the majority of the regional councilors are genuine regional politicians (Kübler 2012).
Mandatory tasks of the Verband Region Stuttgart include key issues such as regional planning, local public transport, regional transport planning, landscape planning, business and tourism development, and certain aspects of waste management. In terms of voluntary activities, the Verband Region Stuttgart can get involved in sport and culture, among others (Growe and Jemming 2019).
The Verband Region Stuttgart derives its income mainly from two sources: levies and other grants (e.g., regional funds for the commuter rail system). The Verband Region Stuttgart has the right to obtain three levies (the general Verband levy from the 179 towns and municipalities, the transport levy from the city and rural districts involved in the integrated public transport network VVS, the waste management levy from the city and rural districts) from the rural districts and from the towns and municipalities in the region. The biggest item in the approximately 300 million Euro budget of the Verband Region Stuttgart consists of the transport sector. This refers to the necessity of funding important parts of the local public transport system, expanding the commuter rail network and, to a lesser extent, also the operation of commuter rail services (Growe and Jemming 2019).
In 1995, the new spatial category of European Metropolitan Regions was introduced, including the EMRS. Although the EMRS had its roots in the Verband Region Stuttgart, today the regions of Heilbronn-Franconia, Neckar-Alb, Northern Black Forest, East Wuerttemberg, Stuttgart, and the state capital Stuttgart belong to it. The EMRS thus extends over an area of 15,400 km 2 and 5.4 million people live in the EMRS-that is almost half (49 percent) of the residents of Baden-Württemberg (Wagner and Growe 2020).
Beyond the Verband Region Stuttgart's borders, close ties to the surrounding regions can be found-for example, to the cities of Heilbronn, Pforzheim, Reutlingen, Tübingen, Schwäbisch Hall, and Schwäbisch Gmünd. The region is closely intertwined with economic relationships, commuter movements, and settlement structures. Due to these close links, the regions of Heilbronn-Franconia, Neckar-Alb, Northern Black Forest, and East Württemberg also belong to the so-called "metropolitan area of integration". Thus, according to the state development plan of 2002, areas in these neighboring regions also belong to the EMRS, of which the Verband Region Stuttgart is the core region in terms of area, population, and economic power. Together, the five regions aim to improve competitiveness through initiating and supporting specific projects.
Joint projects within the metropolitan region should strengthen their position in international competition. Cooperation exists, for example, in biotechnology, regional development, tourism marketing, sports and culture, and especially in transport. For example, the "Metropol Ticket" is not only valid in the Verband Region Stuttgart area but for all transport associations in the metropolitan region. 1 The other region being the Hanover Region. However, the administrative demarcation of the Hanover Region comes close to a county that all have directly elected councils in Germany.
To sum up, the metropolitan region of Stuttgart includes mechanisms of input legitimation-a directly elected assembly in the region of the Verband Region Stuttgart. However, the Verband Region Stuttgart is only one of five regions that build the metropolitan region of Stuttgart. In the other four regions, members of the regional assemblies are elected at the municipal level and delegated to the regional assembly. The example shows that input legitimation through directly elected politicians is a way to increase identification with a new region and to increase the legitimation of planning decisions that might be incompatible with municipal interests (as shown by the Verband Region Stuttgart).

Throughput Legitimation in Spatial Planning in Germany
Throughput legitimation in spatial planning is explained using the example of the building project of the train station Stuttgart 21. Stuttgart 21 is part of the Stuttgart-Ulm rail project and is planned and financed by Deutsche Bahn AG, the federal government, the state of Baden-Württemberg, the Verband Region Stuttgart, the city of Stuttgart, and Stuttgart Airport. The building owner is Deutsche Bahn AG. The project aims to reorganize the railway node in and around Stuttgart. It contains around 60 km of rail routes and three new train stations: Stuttgart Central Station, the Airport/Trade Fair Station, and an additional S-Bahn (local train) station. The main station is to be converted from a terminal station to a lower through station (Nagel and Satoh 2019).
The project has been planned for almost 25 years. Each individual step has undergone a planning process according to the law. The municipal council of the city of Stuttgart, the Verband Region Stuttgart, the state parliament of Baden-Württemberg, and the German Bundestag have approved the project with large majorities. Nevertheless, the project is highly controversial and has divided citizens and politics for many years (Novy and Peters 2012;Nagel and Satoh 2019).
Supporters of the controversial station project argued that a through station offers more capacity to run trains and enables more flexible and faster operation and shorter travel times. In addition, the supporters emphasized the urban development opportunities of the project; by dismantling the above-ground track systems, an area of around one hundred hectares is freed up in the city center, which can be used for apartments, offices, and green spaces. However, the opponents of the project criticized the high costs of over four billion Euros on the one hand and the negative effects of the long-term construction site and the associated tunnel work on the environment and mineral water resources on the other.
Stuttgart 21 not only divided citizens in Stuttgart, but impacted on residents throughout Baden-Württemberg. Supporters and opponents were irreconcilable for a long time. After the demolition work on the north wing of Stuttgart Central Station started in summer 2010, protests against Stuttgart 21 increased. Every week, more than 50,000 participants demonstrated in Stuttgart city center (Novy and Peters 2012). Thus, Stuttgart 21 was no longer just a controversial underground station but had become a crisis of democracy (Novy and Peters 2012). The protests against Stuttgart 21 were seen as a sign to politicians who are increasingly moving away from the everyday reality of their citizens. The conflict over Stuttgart 21 can also be understood as a power struggle between the established party politics of the S 21 supporters and well-organized civil society groups (Nagel and Satoh 2019).
To find a solution to the conflict, an arbitration process was carried out, considered successful by many actors. Arbitration has been described as a new way of communicating between civil society and parliamentary democracy. In spite of the successful arbitration, a referendum on Stuttgart 21 was also carried out in Baden-Württemberg to increase the legitimation for a final decision about stopping or continuing the project.
In the fall of 2010, politicians of the Social Democratic Party saw only one way to pacify the troubled situation: a referendum to decide on Stuttgart 21. However, the referendum was not intended as a vote on the construction project as such but, rather, on the financing of the project, since issues of financing had been seen as the best possibility to intervene. Politicians of the Green Party supported the idea to carry out the referendum, but those from the Christian Democrats (CDU) and the Free Democrats were critical. The then Prime Minister, Stefan Mappus (CDU), thus offered little chance of a possible referendum on Stuttgart 21. "We have a project that has been decided by the municipal council of the City of Stuttgart, the Stuttgart Region, the state parliament and the Bundestag in over 20 years, in which the planning approval process has been completed, has gone through all the legal authorities and where the building built by the client Deutsche Bahn AG has already started, so the client has acquired the right to build, so you can't come after that and say that you have to vote on it again".
However, after new elections in Baden-Württemberg, a new Prime Minister belonging to the Green Party was elected in 2011. The Green Party ruled in a coalition with the Social Democratic Party. On 27 November 2011, Baden-Württemberg voted in a referendum in favor of the underground station Stuttgart 21. This result was recognized by the new government, which originally viewed the construction of Stuttgart 21 critically, and the construction of the station has continued since then (Nagel and Satoh 2019).
Using the train station as an example, it becomes clear that forms of direct democracy can contribute to conflict resolution in political and planning processes, especially if parts of the conflict arise because of alienation between politicians and citizens. However, new questions arise in connection with planning processes, such as the role of referendums in relation to planning considerations and the processes of citizen participation and voting processes that have already been carried out by politically elected bodies. Another aspect is the relation between a referendum and the obligatory participation process. Due to EU-law and German law, public participation is obligatory in planning processes. However, participation processes are often more forms of consultation and do not-unlike a referendum-result in an immediate consequence with regard to the planning project. As planning processes are usually complex, the results of the participation processes become a part of the weighing of interests. Against this background, the question remains if a simple yes/no referendum can replace the complex process of weighing of interests and-if both mechanisms are applied-who is allowed to participate and who is allowed to vote?

Output Legitimation in Spatial Planning in Germany
Output legitimation in spatial planning as part of metropolitan governance is explained using the example of the cross-border metropolitan region Upper Rhine. The Trinational Metropolitan Region Upper Rhine forms a common, European cultural, living and economic region that crosses the borders of Germany, France, and Switzerland. Located at the German-French-Swiss border, the Upper Rhine Region covers 21,000 km 2 and has a population of 5.9 million. Based around five core cities, Karlsruhe (G), Freiburg (G), Strasbourg (F), Mulhouse (F), and Basel (CH), but divided by the River Rhine and language, the region and its people are united by historical ties to Alemannic culture and tradition (Harrison and Growe 2014a).
In 2010, those responsible for cross-border cooperation founded the Trinational Metropolitan Region Upper Rhine. However, the Upper Rhine region looks back on a long and successful tradition of cooperation (Sohn et al. 2009). Signing a government agreement in 1975 marked a milestone in cross-border cooperation in the Upper Rhine region. With this agreement, the formation of a government commission to examine and resolve neighborhood issues in the Upper Rhine region provided the previous developments with an institutional framework. In order to support the government commission in its task, two regional committees for the northern and southern catchment area were set up. For the first time, the German, French, and Swiss governments expressed their common wish to institutionalize cross-border regional cooperation, to maintain regular contacts, and to deal jointly with cross-border issues affecting all partners. The government commission is composed of a German, French, and Swiss government delegation, each with a maximum of eight members, who are appointed by their governments (Fricke 2014).
The work of these committees culminated in 1991 in the Upper Rhine Conference, a collaboration between administrations in the border area. The conference became the central information and coordination organ for cross-border cooperation on the Upper Rhine. Important topics are transport projects, the design and simplification of a cross-border labor market, and the development of strategic ideas for EU projects, as well as the administration of EU funds. The Upper Rhine Conference is thus related to strategic and operational goals. The Upper Rhine Conference is financed through memberships of German, French, and Swiss administrations, and administrative staff work together on sectoral issues. The government commission existing since 1975 is, until today, the contact between the Upper Rhine Conference and the respective national governments with regard to questions that cannot be regulated at a regional level (Schneider-Sliwa 2018).
Since 1997, work of the Upper Rhine Conference has been supplemented by the Upper Rhine Council. The German-French-Swiss Upper Rhine Council is the assembly of politically elected actors in this region. It has 71 members: members of parliament and other elected members (e.g., district councilors, mayors) from four sub-regions. The primary tasks of the Upper Rhine Council are mutual information and political consultation on important issues relating to the region. The council is also a catalyst for new cross-border initiatives. Statements, usually as resolutions, are addressed to regional and national governments, the Upper Rhine Conference, the European Union, and others (Schneider-Sliwa 2018).
Since 2010, the Trinational Metropolitan Region Upper Rhine has been the roof of the two institutions. Like all border regions, the Upper Rhine faces new challenges in the area of cooperation. Like other functional border areas, the region goes beyond the administrative districts and territorial responsibilities of the existing institutions. The aim of the Trinational Metropolitan Region Upper Rhine is to find an innovative governance model that is able to open up to any partnership (Fricke 2016).
The development of governance structures in the region shows, on the one hand, that technical questions can initiate the development of a political exchange in unusual regions. Following administrative exchange, an arena for political exchange was created in order to increase political legitimacy through politically elected representatives. However, these representatives are not directly elected by the population in the whole cross-border region. The members of the Upper Rhine Council are elected local or regional politicians and are delegated to be also members of the Upper Rhine Council. On the other hand, the example of the Metropolitan Region Upper Rhine shows that governance structures for unusual areas (which correspond to non-established administrative boundaries) have to cope with one additional challenge: they must lobby for their region to be perceived and accepted by actors at established governance levels, for example in the national states (Harrison and Growe 2014a).
In the case of the Upper Rhine metropolitan region, this succeeded in terms of planning questions by establishing the Upper Rhine region as a cross-border metropolitan region. In German politics, the concept of metropolitan regions has been rated as a success (Sinz 2005, p. i). The concept succeeded in initiating regional governance activities by means of only a strategic mission statement and without funding. The successful initiatives of other regions (Bremen/Oldenburg, Hanover-Braunschweig-Göttingen, Nuremberg, and Rhein-Neckar) to be recognized as metropolitan regions are also part of this success (Growe 2018).
However, the recognition of metropolitan regions as "metropolitanization from above" (Blotevogel 2000, p. 164) is also viewed critically, since discrepancies between desire and reality in fulfilling the role as functional important "engines of development" may lead to the discrediting of a metropolitan-orientated policy. Still, the efforts to be recognized as metropolitan regions can also mobilize actors in the regions and promote cooperation, which is another measure of success.
The critique was raised mainly from two groups: rural actors and actors from cross-border regions that had not been taken into consideration in developing the concept of metropolitan regions in Germany. In response to the unanswered questions, model projects in spatial planning for "large-scale communities of responsibility" (Hesse and Leick 2013, p. 353) and cross-border metropolitan regions have been introduced in Germany. These model projects are working on the further development of the concept of metropolitan regions so that, in the updated guidelines adopted in 2016, four new cross-border metropolitan regions (Lake Constance, Meuse-Rhine, Upper Rhine, Greater Region Saar-Lor-Lux) and the metropolitan regions previously recognized are presented (Growe 2018).
This designation solidifies the perception and recognition of the unusual cross-border metropolitan region of the Upper Rhine in Germany.
To sum up, a development can be observed in the Upper Rhine region from a specialist exchange in administrations to the establishment of a political exchange, and to the development of metropolitan governance structures and work on the recognition of the region in the respective nation states. Thus, successful and output orientated work has legitimized increased engagement and helped to acknowledge, confirm, and deepen governance mechanisms in the region (Harrison and Growe 2014b).

English Planning Background
Unlike Germany and its federally organized nature, England has a highly centralized government structure with currently no formal levels of democratically elected government above that of the local authority scale. The local authorities themselves are a directly elected tier of municipal and district government, with local government elections held regularly to elect local politicians (Councilors), the majority of whom are associated with a particular political party. These authorities are responsible for preparing the development plans for their areas (local plans) and for determining planning applications. Unlike many planning systems in Europe and elsewhere, these local plans are not actually legally binding, and planning authorities can approve proposals that do not always accord with their own plan, although they are required to take account of it as well as any other material considerations. Thus, the English system is often described as "discretionary", in contrast to the more mandatory systems associated with the preparation of legally binding forms of zoning plans.
In contrast to Germany's strong tradition of regional spatial planning, attempts to introduce regional planning in England have come and gone over the years and have never involved directly elected tiers of regional governance (Glasson and Marshall 2007). The most recent period of regional planning started in the early 1990s but gained greater impetus in the reforms undertaken by the Labour Governments of the late 1990s and 2000s. Reforms to the planning system in the mid-2000s introduced a new form of Regional Spatial Strategies (RSS), which were enhanced versions of earlier Regional Planning Guidance (RPG) published by the Secretary of State. The RSS were subject to a more formal preparation process, including a public examination chaired by an independent inspector, and were accorded enhanced statutory status as a part of the formal "statutory development plan" alongside new forms of Local Development Frameworks (LDFs) at the local authority level.
However, the Regional Assemblies (RAs) or equivalent bodies charged with preparing the RSS were not directly elected but, rather, were made up of representatives from the local authorities within each region together with some additional stakeholder representatives from business, education, and other sectors. An initiative to introduce directly elected regional assemblies was dependent upon regionally based referenda of the local population to agree to the introduction of a new regional tier of government. However, this failed with a resounding "no" vote in the first region to vote (North East England). No further regional referenda were carried out, and the idea of directly elected regional government was dropped. RSS were, however, prepared for all English regions by the time the Labour government lost power in 2010. Alongside them, Regional Development Agencies (RDAs), whose boards were directly appointed by central government, were responsible for preparing non-spatial regional economic strategies, which sometimes caused some tensions in outlook with the RAs/RSS. By the late 2000s, further re-organization of regional governance saw moves to prepare new forms of integrated regional strategies, led by the RDAs, in an attempt for the better co-ordination of spatial and economic planning. However, the Conservative-dominated coalition government elected in 2010 abolished all aspects of regional planning and governance, including both the successor bodies of the RAs and the RDAs. Thus, with the passing of the Localism Act 2011, the most recent era of English regionalism came to an abrupt end, to be replaced by a shift towards "localism" with its greater emphasis on devolved powers and responsibilities to local authorities and also to communities lower down the spatial hierarchy, including new Neighbourhood Plans (Ziafati Bafarasat and Baker 2016b).
Since 2010, therefore, in terms of spatial planning there emerged a significant strategic gap between national planning policy and the local authority level (Baker and Wong 2013). At the national level, such policy is essentially non-spatial in nature and set out in the National Planning Policy Framework (NPPF), initially published in 2012 and revised in 2018 (CLG 2012(CLG , 2018). At the local authority level, the latest versions of authority-wide statutory development plans now generally take the form of Local Plans. In order to address strategic planning issues and cross-boundary aspects, the government introduced a statutory "duty to cooperate" to local authorities when preparing their individual local plans. However, the effectiveness of this has proved rather mixed, being dependent on the willingness of neighboring authorities to genuinely work with each other and wider stakeholders (Boddy and Hickman 2013).

Metropolitan Regions and Metropolitan Governance in England
These shifts towards localism had a particular impact on spatial planning at the metropolitan scale. In some of the more rural ("shire") areas of England, there remains a two-tier structure of local government, consisting of larger County Councils and smaller District Councils. However, the County Councils today play little role in spatial planning apart from the specialist areas of minerals, waste, and transport. Until relatively recently, local government structures in England within the larger urban conurbations or metropolitan areas primarily consisted of metropolitan authorities, none of which were of the scale and size of an entire metropolitan area or conurbation as a whole. Thus, for example, the area known as Greater Manchester consists of ten city and metropolitan borough councils (such as Manchester City Council), and there are 32 London Boroughs that cover the area of Greater London.
For a brief period in the 1970s and 1980s, this wasn't the case. The 1972 Local Government Act introduced a two-tier local government system for the whole of England and, in the larger metropolitan areas, this involved the creation in 1974 of five Metropolitan County Councils (such as the former Greater Manchester County Council). All County Councils, in both the shires and the metropolitan areas, were responsible for preparing a strategic level of planning policy known as a structure plan. Associated local plans at the district or borough level were optional and required conforming to the structure plan, ensuring strategic oversight of local decisions and policy development. However, these relatively large and powerful metropolitan county councils were predominantly controlled by the Labour party in contrast to the Margaret Thatcher-led Conservative central government from 1979, and tensions between metropolitan and central policy were rife. The outcome was the abolition of the Metropolitan County Councils (and the associated Greater London Council) in 1986. The lower-tier metropolitan borough councils thus became single-tier unitary authorities. The only mechanisms for metropolitan-wide coordination and strategy formulation was, therefore, voluntary attempts in some areas to preserve joint working arrangements with respect to agreed issues and a very limited introduction, in terms of spatial planning, of statements of strategic planning guidance. These were issued by the central government to give some steer to the preparation of new unitary development plans at the local authority level (Thomas and Roberts 2000).
During the 1990s and 2000s, there were a number of ad hoc initiatives and developments that re-introduced an element of metropolitan governance, rather than government, in England's larger cities. These included those areas where some continued formal joint working was established-such as the Association of Greater Manchester Authorities (AGMA). More broadly, there was also a growing interest in the concept of city regions that became more prevalent during the years of the Labour governments up to 2010. A number of initiatives and bodies emerged, including city region-based passenger transport executives for Greater Manchester and elsewhere; the establishment of core cities groups; and a number of city region studies published by OECD and the Centre for Cities that proposed a devolution of powers to two city regions (Greater Manchester and Birmingham) with potentially five others to follow in Liverpool, Newcastle, Leeds, Sheffield, and Bristol (Marshall and Finch 2006). In 1997, the Labour government initiated a review of sub-national economic development and regeneration. The subsequent Local Democracy, Economic Development, and Construction Act 2009 provided a statutory framework for more city region activity, including powers relating to transport, skills, planning, economic development, and city region deals and, significantly, allowing the potential creation of combined authorities.
Although the incoming coalition government in 2010 were quick to abolish the inherited regional tier of governance and strategies, they more readily took on board the city region concept including that of combined authorities. These included the first Combined Authority for Greater Manchester in 2010 and the associated introduction of city region Mayors. London had already seen the re-introduction of a conurbation-scale authority ten years earlier with the establishment of the Greater London Authority (GLA) in 2000 and the election of Ken Livingston as its first mayor (later to be replaced by Boris Johnson). This move also saw the introduction of a conurbation-wide London Plan to coordinate spatial planning policy across the capital region and steer the unitary London Borough's development plans. More recently, there has been an increase in the number of neighboring local authorities working together on a range of joint statutory or non-statutory spatial strategies in other parts of the country outside London (Riddell 2019). One such example is the Greater Manchester Spatial Framework (GMSF), jointly prepared by Greater Manchester Combined Authority (GMCA) and the metropolitan and city authorities that make up the area.
However, there remains no formal definition of a metropolitan area in England. Outside London, all of the former metropolitan county council areas are now designated as, or form the core of, statutory Combined Authorities: West Midlands (including Birmingham), Greater Manchester, Liverpool City Region, North of Tyne (including Newcastle upon Tyne), North East (including Sunderland), Sheffield City Region, and West Yorkshire (including Leeds). There are also combined authorities in Tees Valley (including Middlesbrough) and the West of England (including Bristol). Most (but not all) of these now also have a directly elected mayor. These cover the vast majority of the largest urban areas in England, although there some others without CA status-for example, South Hampshire (including Southampton and Portsmouth), Nottingham-Derby, Leicester/Leicestershire (a proposed CA), Brighton and Hove, and Bournemouth-Poole. Together, these cover the largest of the 46 Morphological Urban Areas (MUAs) defined in the EU's ESPON (European Spatial Planning Observation Network) project with populations around or above 500,000 (ESPON 2007). Some of the aforementioned combined authority areas, notably in the North East of England, are focused on an urbanized metropolitan area but also extend well beyond this to wider rural hinterlands for administrative purposes. There are also some combined authority areas in more mixed urban-rural locations that would not generally be considered as metropolitan, such as the Cambridgeshire/Peterborough CA and the proposed CAs in Devon/Somerset and Cheshire/Warrington. This lack of a simple and stable governance structure and formal spatial planning system for the metropolitan areas of England means that attempts and approaches to spatial planning at the metropolitan or city region scale have been equally varied, both geographically and over time. Not surprisingly, the associated mechanisms to legitimize such approaches can also be seen to have varied considerably and have often proved challenging, as seen in the examples that follow which focus particularly on experiences in Greater Manchester and the wider North West region of England.

Input Legitimation in Spatial Planning in England
Given the general lack of a specific tier of metropolitan governance in England, it is not surprising that examples of the legitimacy of spatial planning through direct democratic election and majority rule at the metropolitan or conurbation-wide scale are limited. Historically, the one time this happened occurred between 1974 and 1986 in the five largest metropolitan areas which had directly elected metropolitan county councils: Greater Manchester, Tyne and Wear, West Midlands, West Yorkshire, and South Yorkshire. Like other upper-tier county councils in England, these were responsible for preparing strategic planning documents, called county structure plans, for their areas as well as having strategic powers and responsibilities over transport, waste disposal, and emergency services. The structure plan set out broad strategic planning policies for the future development of land in their areas over a period of around 20 years, including the strategic general allocations of land for housing and industry as well as policies and general designations for environmental protection and the scale and distribution of development. The lower-tier local authorities had to take account of the structure plan when making strategic planning decisions and preparing their own local plans. The actions and performance of the metropolitan county councils were directly accountable, in due course, through the ballot box. In addition, the preparation of structure plans was (like other statutory planning documents) subject to rigorous opportunities for public participation and independent testing at an examination in public and had to be approved by the Secretary of State (central government) before implementation. All the metropolitan counties produced structure plans during their relatively short lifespan, with the example for Greater Manchester being approved in 1981.
In more recent times, the most obvious example of a conurbation-wide spatial plan that was linked to direct elections by the public was that of the London Plan. This has been issued by the directly elected London Mayor and published by the Greater London Authority since its creation in 2000. The original London Plan was published in 2004 and set out a statutory spatial development strategy for Greater London that must be taken into account by individual London boroughs when preparing their own spatial plans for their areas. There have subsequently been a number of revisions, including a new plan published by Boris Johnson as the then London Mayor in 2011. The current version of the plan was published in 2016, but a new plan authorized by the current London Mayor, Sadiq Khan, has passed through most of its consultation and examination processes and is expected to be published in due course.
However, outside London one of the most significant developments in the last decade has been the establishment of statutory metropolitan-scale upper-tier authorities in a number of areas in the form of the new Combined Authorities. These now cover almost all of the largest metropolitan areas in England and most are now also associated with the direct election of a mayor, since doing so allows for the greater devolution of power and associated resources from Whitehall (central government). Although these CAs generally have powers associated with housing, transport, planning, and policing, not all have yet attempted to prepare a spatial planning document for their area. The most notable exception so far has been that of the Greater Manchester Combined Authority, which has been working on a metropolitan-wide spatial framework in association with the ten local authorities in Greater Manchester. This strategy development process was, however, interrupted by the direct election of the first Mayor of Greater Manchester in 2017.
The Greater Manchester Combined Authority (GMCA) was established in 2011 as a strategic tier of local governance for the metropolitan area of Greater Manchester. Consisting initially of indirectly elected local politicians from each of the 10 city and district councils within the conurbation, it is a strategic authority with powers encompassing public transport, housing, regeneration, skills, carbon neutrality, and planning. GMCA published the first draft of its authority-wide Greater Manchester Strategic Framework (GMSF) in 2016 (GMCA 2016). Some of its proposals, particularly involving designated growth areas and associated green belt development, generated a significant amount of local community opposition. However, this point in the process coincided with central government making provision for a directly elected Mayor for Greater Manchester. During the subsequent mayoral election campaign, the to-be-elected candidate, Andy Burnham, actively campaigned on a pledge to radically revise the GMSF with a shift towards more brownfield development and a "no net loss" of the green belt. The subsequent revised draft spatial framework (GMCA 2019) was published for consultation in 2019. It doesn't achieve the "no net loss" of the greenbelt pledge but does cut the loss of the green belt by around half of its original extent. Subsequently, there has been no further developments, and it has been suggested that this reflects both arguments between the GMCA and central government over the amount of new housing required as well as politically sensitive local council elections in some of the districts where the most controversial green belt sites were still proposed (Manchester Evening News 2020).
The experience so far of GMCA in developing its GMSF highlights the complexity of spatial planning at the metropolitan scale, including balancing the requirement to meet national policy as well as keeping the local electorate happy. In this case, the introduction of a directly elected mayor clearly resulted in an immediate change in direction. Further electoral issues may well continue to shape or frustrate the completion of the strategic plan, demonstrating the potential power of input legitimization.

Throughput Legitimation in Spatial Planning in England
If the GMSF shows the potential of the population to influence spatial planning outcomes indirectly via the election of a mayor, it also shows how public participation in the planning process can have a major effect too. Indeed, the statutory planning system in England requires a rigorous process of public consultation and independent testing of the emerging plans. The extent of local campaigning against the proposed green belt deletions in the GMSF shows that the public can get very interested and their objections have considerable impact where strategic issues impact upon and threaten local environments. In many ways, therefore, throughput legitimation in the form of public and stakeholder participation in the plan-making process is the cornerstone of legitimizing the emerging spatial planning policies and proposals in England. However, the Greater Manchester area also provides a comparatively rare example of a direct referendum of the local population to determine a spatial planning-related policy direction at the metropolitan scale.
A couple of years prior to the creation of the GMCA, the Greater Manchester authorities and the Greater Manchester Passenger Transport Executive proposed to introduce a congestion charge for Greater Manchester as part of a Transport Innovation Fund bid to central government for a £2.7 billion package of public transport measures. These included extensions to the existing Manchester Metrolink tram system. Vehicles entering the metropolitan area within the M60 would be charged in the morning commuter peak alongside an additional charge for those entering the central area (inner cordon). The congestion charge was intended to help repay a loan as part of the overall package over a 30-year period. Eight of the 10 local authorities within the Association of Greater Manchester Authorities (AGMA) initially supported the proposal, although another withdrew their support later. Rival groups of businesses and other stakeholders set up campaigns for (e.g., United City, Clean Air Now) and against (e.g., Greater Manchester Momentum Group, Manchester Against Road Tolls (MART)) the proposal, and a Stop the Charge Coalition was formed that included leaders of the opposing authorities as well as several local MPs. MART started a legal petition for a referendum which gained momentum and, in July 2008, AGMA agreed to put the proposals, including the congestion charge, to a referendum of the local population.
The results of the referendum in December 2008 showed that 79% voted against the plan and only 21% in favor of it, with a turnout of over one million voters, representing around 53% of the potential electorate. That was the end of the proposed congestion charge although, interestingly, government funding was nevertheless later secured for the proposed Metrolink extensions and other improvements anyway. More recently, Transport for Greater Manchester (TMG) have raised the possibility of introducing Clean Air Zones in the city, whereby drivers of polluting vehicles would be charged for entry. However, this has not been introduced yet and legal responsibility would lie with individual councils rather than the mayor. It should perhaps also be noted that the earlier London Congestion Charge, introduced in the capital city in 2003, was never subject to such an explicit referendum of London residents, although the introduction of a road-charging scheme was a central part of Ken Livingston's successful campaign to become the first mayor of the newly created GLA in 2000. In this sense, the London charge arguably gained its legitimacy via input mechanisms, whereas the Greater Manchester congestion charge failed to gain throughput legitimation.

Output Legitimation in Spatial Planning in England
Perhaps the biggest issue affecting the legitimacy of all strategic-level administration and associated spatial planning revolves around the identity of the area in question. Most people have a very strong local identity associated with their neighborhood, district, or town. However, generating identity with a more abstract strategic area is more problematic. This was apparent in the aborted plans for directly elected regional assemblies and was previously also an issue for the elected metropolitan county councils (1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986). Although their demise at the hands of Margaret Thatcher's government had much to do with local/national political tensions, many local residents in these areas were, at best, ambivalent about their existence in the first place. Thus, the charges of inefficiency and overspending may have reduced their legitimacy as a crucial provider of strategic services but, in truth, many local residents maybe never really identified with or saw themselves as part of the metropolitan area. To take Greater Manchester as the example, there are many local residents living within the ten local authorities that comprise the current Greater Manchester CA that still see themselves as part of the surrounding shire counties rather than the conurbation. Thus, the relatively affluent residents of places like Altrincham, Hale, and Bramhall may well associate themselves with Cheshire rather than with the metropolitan boroughs of Trafford or Stockport in which they are actually located. Similarly, in the north of the conurbation, residents of the more freestanding towns beyond the M60 such as Wigan and Bolton may associate themselves more with Lancashire. However, from a socio-economic and functional perspective, others have argued that the current Greater Manchester area is too constrained (Baker and Wong 2013), given the exclusion of swathes of northern Cheshire whose affluent residents frequently commute into the city center to work and for shopping and leisure. The definition or identification of the boundaries of a metropolitan area is therefore often both complex and contested.
If gaining legitimacy at the strategic scale is hard even for formal tiers of government, it follows that establishing an identity for a more ad hoc city region or metropolitan sub-region is likely to be even more challenging. Nevertheless, the North West also provides a couple of interesting developments in identity-building within both the environmental and economic spheres. These both cover a wider area than just Greater Manchester, embracing what has long been known as the Mersey Belt, dating back at least to its inclusion in the Strategic Plan for the North West in 1974 (North West Joint Planning Team 1974). This area includes the city regions of both Liverpool and Greater Manchester despite the traditional rivalry between these cities. Thus, both can be seen as examples of more flexible "soft spaces" for strategy formulation, future development, and identity creation and, indeed, as part of a wider formation of a longstanding "spatial imaginary" in this part of North West England (Deas et al. 2015).
The Mersey Basin Campaign was established in 1985 as a 25-year response to dealing with environmental degradation and industrial contamination in the River Mersey catchment area stretching from the Mersey Estuary to Greater Manchester. Notably, it was based on river catchment boundaries rather than existing political or administrative boundaries. It had a very wide stakeholder membership, an involved network of over 20 action partnerships, and was funded by the EU and central government as well as other sources. Although focused on the environment, it was more than just a river "clean-up" campaign and gained wide-ranging support across business and economic stakeholders as well as the more obvious environmental organizations and local authorities. This was partly because it was difficult to argue with the concept of environmental improvement but also because it went successfully put across the message that such improvements were crucial to the future economic prospects of the sub-region, including attracting new investment. The campaign was awarded the inaugural International Theiss Riverprize in Brisbane, Australia, in 1999 in recognition of its role in transforming the environment of the region. When it was wound up as originally planned in 2010, the River Mersey was said to be cleaner than at any time since before the industrial revolution (Mersey Basin Campaign 2020). Some of its work is, today, continued by a charity, the Mersey Rivers Trust, which continues to work with local authorities and other stakeholders in the region. The campaign is thus a good example of an ad hoc, non-administrative/government body that nevertheless gained wide-ranging output legitimacy through what it achieved and delivered.
One of the major private sector organizations that was involved in the Mersey Basin Campaign was Peel Holdings (now Peel Group). Peel Holdings are a major player and landowner in the Mersey Belt thanks to their considerable strategic landholdings. These include the Manchester Ship Canal, which was originally opened in 1894 to provide a navigable link between Manchester/Salford and the Mersey Estuary and allowed for the subsequent development of Salford Quays as a major port dock despite being nearly 40 miles inland. The canal runs parallel to the River Mersey through the sub-region and Peel Holdings are majority owners of the Canal and also own, or have owned, many strategic sites from the Estuary (e.g., Wirral Waters) through to Salford Quays. These include Media City UK and the Trafford Centre (a regional retail development), which was the UK's largest ever property acquisition when Peel sold it to Intu Properties in 2011. Unusually for a private sector developer, the Peel Group were responsible for the promotion of another strategic spatial development strategy in the Mersey Belt, this time focused on economic development and initially called "Ocean Gateway" but later renamed "Atlantic Gateway". Their intention was to use Ocean Gateway as an "economic powerhouse" to "enhance, strengthen and bring together the Liverpool and Manchester City Regions" (Peel Group 2009, p. 2).
The spatial or territorial extent of the Ocean/Atlantic gateway has been described as "imprecise and loosely articulated", reflecting "the stress placed on functionality and the conspicuous absence of administrative boundaries" (Deas et al. 2015, p. 11). The catalyst for Peel's initial launch of Ocean Gateway was undoubtedly their corporate interests in gaining further support and backing from government and other investors for the suite of major developments they were already planning. However, concerns over the legitimacy of such a private sector initiative in terms of democratic accountability led to the involvement and incorporation of the idea by the North West Development Agency (NWDA) into its last regional strategy, just before the NWDA was itself abolished in 2010. The Peel Group subsequently went as far as to submit Atlantic Gateway as a bid for Local Enterprise Partnership (LEP) status that year. However, this generated significant opposition from the local authority-based partnerships in the region and was ultimately unsuccessful with separate LEPs ultimately established for the Liverpool City Region and Greater Manchester. Subsequent academic research (Deas et al. 2015) reveals mixed feelings about the Atlantic Gateway, especially in the public sector, albeit with greater support at the Liverpool end of the corridor. Its democratic legitimacy continues to raise questions, but it has arguably provided strategic thinking and encouragement to national government and the private sector to invest in further infrastructure and economic development in the Mersey Belt.

Input Legitimation or the "Who" Aspect
As directly elected metropolitan councils are rather an exception than the norm in both Germany and England, the input legitimation of spatial planning at this level has been limited to such cases as the Verband Region Stuttgart, the Greater London Assembly, and the Greater Manchester Mayor. The Verband Region Stuttgart has drawn on input legitimation to incorporate in spatial planning a broad range of issues-including transport, landscape, business, and waste management-and to embark on cross-metropolitan cooperation in biotechnology, regional development, tourism marketing, sports and culture and, especially, in transport. The English examples, in particular the case of Greater Manchester Mayor, have experienced mixed outcomes-for example, in terms of housing and greenbelt policies that did not subsequently gain significant throughput legitimation. After elaborate throughput legitimation in terms of consultation and examination processes, the Greater London Assembly managed to put in place spatial (London Plan) and other policies such as the Congestion Charge through its input legitimation, but all these English examples attach considerable throughput legitimation to input legitimation, indicating the latter is precarious in nature. In other words, the "who" aspect of legitimation has been reliant on the "how" aspect (Table 2).

Throughput Legitimation or the "How" Aspect
This might indicate that, in England, the throughput legitimation of metropolitan spatial planning is more effective. However, the Greater Manchester Congestion Charge referendum and, to a lesser extent, the Greater Manchester Strategic Framework consultation expose the difficulties of spatial planning at the metropolitan level to safeguard infrastructure development, regeneration, and environmental protection with direct reference to public opinion. In Germany, however, the referendum for the underground station Stuttgart 21 indicates how barriers to the input legitimation of spatial planning can be overcome by throughput legitimation, although potential frictions between the two are brought to the fore in this example (Table 2). In both national contexts, throughput legitimation from the referendum route has been taken where the classic planning consultation routes have been viewed with pessimism by the public in relation to the weighting of their local views alongside other considerations and, in particular, political inputs. It is interesting, however, that in the case of Stuttgart 21 throughput legitimation via a referendum became necessary in spite of input legitimation by the municipal council of the City of Stuttgart, the Stuttgart region, the state parliament, and the Bundestag. One possible explanation might be termed as "fuzzy" political decision-making by directly elected politicians, occurring in systems of multi-level or cross-boundary governance which obscure the input of the directly elected bodies in relation to other players.

Output Legitimation or the "What" Aspect
From the examples discussed above, output legitimation, which reflects the "what" aspect, is arguably the most effective of the three legitimation types in metropolitan spatial planning in Germany and England. In both contexts, it has enabled the engagement of multiple stakeholders and actors and the addressing of various topics in spatial planning with a minimum of public objection. This has sometimes resulted in more informal, rather than formal, spatial plans. However, their impact, either through bridging administrative and then political institutions or through discourse steering of subsequent formal spatial plans, can be considerable. In the cross-border metropolitan region of the Upper Rhine, promises of common economic growth and identity cohesion enabled cross-country spatial planning that ultimately extended to several sectors and proliferated a range of representative institutions with complex horizontal and vertical links ( Table 2). The exercise even managed to engage rural actors in a subsequent geographical extension of its metropolitan discourse, marking a unique example of unusual spatial planning by consensus. In England, the case of Mersey Basin Campaign signifies a similar effectiveness of output legitimation, which owes legitimation to the recognition of achievements as well as uncontested nature of its activities-i.e., environmental protection. This has enabled an informal third-sector metropolitan entity to engage with local authorities in metropolitan spatial planning at the intersection of several sectors, including transport, housing, and economic development, without public objection. The case of Atlantic Gateway, however, has seen some difficulties in output legitimation due to the dominance of private organizations in institutional arrangements, which, although enabling an innovative geographical extension of spatial planning initiatives, has increasingly conducted its joint economic development works with localities, quangos, and central departments vulnerable to potential local civic objection.

Legitimation of Planning Processes as Part of Metropolitan Governance
Spatial planning is a main mechanism by which metropolitan governance goes beyond sectoral and often siloed service provision and steers neighbouring autonomous localities towards an agreed collective desired future at the metropolitan scale. It is, therefore, the "where", "how", and "what" of legitimation in the metropolitan governance intersect, reflecting significant challenges and opportunities for learning and the application of what we termed input legitimation, throughput legitimation, and output legitimation. Input legitimation poses no serious challenge to the design of spatial planning per se as but rather to the institutional design of metropolitan governance, as it has proved difficult, at least in Germany and England, to overcome local and central opposition to the establishment of directly elected metropolitan or regional government. Issues of increasing distrust in political institutions, coupled with the relatively unsuccessful experiences of Germany and England in exceptional cases where metropolitan/regional governments have been established, suggest that even where input legitimation is available, throughput legitimation and, to a lesser extent, output legitimation are also deemed necessary. However, often a lack of input legitimation also complicates the throughput and output legitimation of spatial planning. Although the more frequent use of referenda and more direct forms of public consultation in planning decisions can be seen as positive in terms of collaborative metropolitan governance, our observations in Germany and England indicate they involve procedural complications and difficulties in making meta-decisions on whether and how to reconcile different political views via either yes/no referendums or more classic planning consultation processes in decision making.
Throughput legitimation has always been an important joint field of investigation between political science and planning in metropolitan studies and remains so in the light of the increasingly connected urban world in which we live and in the light of a more active civic society that seeks to be part of decision making at the intersection of administrative divisions. Whilst referenda, as observed in our investigations, often bring quick resolutions to enduring conflicts that have not been overcome otherwise, spatial planning is a highly diverse and broad activity that tends to involve throughput legitimation in multiple stages and across different topics, sometimes involving front-loading engagement on the visions and objectives of the exercise as well as the final outcomes. This ideal of throughput legitimation, however, faces implementation challenges as, for example, seen in England where governmental and more organized interests were observed to confine front loading engagement to a small circle of actors working in concert with central policy directives in a regional political vacuum.
Does this mean that the throughput legitimation of spatial planning could only be optimized in combination with input legitimation by a directly elected regional government? As a hypothesis, maybe yes. We argue that stakeholders that represent a region through local political nomination, multi-level governance, or other organized interest claims are not holistic representatives of all the affected groups and communities and therefore have reason to seek to confine real civic input in throughput legitimation. Whilst they do depend on a degree of throughput legitimation to present their priorities as those of the public, they can attempt to hollow out more collaborative exercises by only leaving minor implementation choices to consultation (but not the overarching objectives or visions which have already been set) or by limiting consultation outcomes in relation with other inputs-for example, "evidence-based" ones which are given higher decision-making weight.
Does such a combination of input and throughput legitimation reduce the scope for output legitimation and thus constrain the inter-sectoral and inter-territorial innovations of spatial planning, including new institution building initiatives and environmental ventures? The experience of Germany as discussed in this paper dismisses this concern. Output legitimation is a main preoccupation of metropolitan spatial planning that is not only aimed the public but also other political, private, and third-sector interests at various scales, even an international audience. However, the emphasis on output legitimation or "what works" has sometimes been sufficiently strong to cause some concerns in both Germany and England about potential discourse manipulation and fabrication in metropolitan governance-for example, in favor of major economic interests. On the other hand, output legitimation in these countries has often been quite sensitive to public views and, for example, in Germany, has responded by making adjustments to spatial planning exercises and discourses.
However, two open questions remain: Firstly, how can metropolitan governance with its multiplicity of tasks (amongst which spatial planning is just one, albeit with something of a coordinating role) deal with crucial future challenges such as climate change, sustainable development, and the health and well-being of the population, to name the currently most obvious. Secondly, what forms of legitimation allow the actors involved in metropolitan governance to decide on priorities and actions? The case studies have shown that although legitimation is necessary, this is not easily achieved in the fuzzy world of metropolitan governance-sandwiched as it usually is between local and national interests. Legitimation also needs the populations' acceptance as well as that of participating organizations and stakeholders, and the challenges need to be understood as reasonable causes for decisions and actions at the metropolitan level. Furthermore, national actors need to acknowledge the importance of the intermediate level of metropolitan regions as a crucial level in a multi-level governance system: metropolitan governance can be close enough to local specifics whilst, through its embracing and coordination of multiple units of local government, still provide a strategic perspective beyond parochial thinking. Despite some successes documented here, metropolitan level governance in both England and Germany has had a rather checkered history and has seldom been a fully accepted tier of directly elected government. Nevertheless, given the immense challenges associated with large urban areas, it has the potential to perform a great deal more in the coming decades in terms of both necessary coordination and as a bulwark between an increasingly top-down government at the central state level and local level government and communities.
In sum, our identification and analysis of the three types of legitimation that metropolitan governance utilizes in the course of spatial planning add conceptual, methodological, and contextual values to the field of study. In conceptual terms, our typology structures fuzzy lines of legitimation across the three ("how", "who", and "what") aspects of metropolitan governance suggested for consideration in the literature. From this point, we undertook an in-depth and cross-case review of variables involved in the design, application, and outcome of input, throughput, and output legitimation using examples drawn from Germany and England. This relational methodology helped us to learn about the contextual dynamics of how the three types of legitimation might reinforce one another in different international settings, leading to the overall conclusion that they work best in combination, although output legitimation has a distinctive capacity to work in less formalized settings.