Conference aim

Conference report

Programme

Workshop programme
Including
Full papers

Information

Eura

See paper in pdf

Metropolitan governance and the "democratic deficit"

Theoretical issues and empirical findings

Brigitte SCHWAB * & Daniel KÜBLER**

* Research fellow at the Institut de recherche sur l'environnement construit
(Swiss Federal Institute of Technology Lausanne) and
Teaching assistant at the Institut d'études politiques et internationales (University of Lausanne)

** Research fellow at the Institut de recherche sur l'environnement construit
(Swiss Federal Institute of Technology Lausanne) and at the
Institut für Politikwissenschaft (University of Zurich)

Paper presented ath the conference Area-based initiatives in contemporary urban policy, Danish Building and Urban Research and European Urban Research Association,
Copenhagen 17 - 19 May 2001

Contact:
Brigitte Schwab / Dr. Daniel Kübler
Institut der recherche sur l'environnement construit (IREC-DA)
Swiss Federal Institute of Technology
P.O. Box 555
CH 1001 Lausanne
Switzerland
Phone : +41 21 693 42 10
Fax : +41 21 693 38 40
e-mail: Brigitte.Schwab@epfl.ch or Daniel.Kuebler@epfl.ch

Abstract

The topic of this paper relates to the more general debate about policy-making in institutionally fragmented urban areas. As a consequence of slowed-down or failed attempts to set up more encompassing metropolitan governments, policy co-ordination in metropolitan areas is nowadays mostly achieved through issue-based and more or less formalised mechanisms of co-ordination - best described by the term of metropolitan governance. The aim of the paper is to address the issue of the "democratic deficit" that many scholars associate to these recent developments in metropolitan governance. It will consist of three sections:

First, we will present the current structural characteristics of metropolitan areas in Switzerland, as well as the issues and challenges that metropolitan policy-making faces in this country today. In part two, we will address theoretical and conceptual issues of our study. We formulate an attempt of specifying the somewhat fuzzy term of governance, in order to define operational categories for an empirical inquiry. We propose to think of governance in terms of a continuum government - governance, and suggest some definitional elements we would retain to do so. We then address the question of democracy and legitimacy with respect to governance on a conceptual level. More precisely, we propose four working hypotheses with which the issue can be empirically examined. In the third section we use this framework to analyse some of the major mechanisms of policy coordination in four urban areas in Switzerland (Zurich, Lausanne-Geneva, Berne and Ticino). The conclusion discusses the results of this analysis, as well as the usefulness of the conceptual framework.

Introduction: metropolitan governance and democracy¹

In most western industrialised countries, cities can no longer be considered in the traditional (weberian) sense of territorially integrated socio-political entities. Rather, they must be considered 'metropolitan areas', that is multi-centered urban regions created by the social, economic and cultural dynamics of the modern societies, and held together by powerful systems of communication and mass transportation. Metropolitan areas are urban systems that extend and develop mainly along functional social and economic networks, more or less independently from institutional boundaries of communities or other sub-national entities. Therefore, metropolitan areas are more often than not characterised by a high degree of 'governmental fragmentation' . Their territory is split up into multiple political and administrative units (communes, cantons, provinces), leading to difficulties regarding the management of metropolitan problems. There have been attempts, throughout Europe, at institutional reforms to create encompassing political institutions for metropolitan areas (e.g. the 'urban province' concept of Amsterdam and Rotterdam, the merger of Berlin and Brandenburg, etc.). However, these reforms have repeatedly failed. In spite of multiple initiatives in many countries, 'metropolitan governments' are exceptional .

1) This paper draws on an ongoing research project funded by the Swiss National Science Foundation in the context of its Priority Programme "Switzerland Towards the Future" conducted at the Institut de recherche sur l'environnement construit of the Swiss Federal Institute of Technology Lausanne (Grant number #5004-058522/1) on the issue of democratic legitimacy with respect to recent developments in metropolitan governance.

The absence of encompassing metropolitan institutions has however not prevented metropolitan policy making beyond the boundaries of single municipalities. As a consequence of the slowed-down or failed attempts of setting up encompassing metropolitan institutions, policy co-ordination in metropolitan areas is nowadays mostly achieved through issue-based, network-like mechanisms of co-ordination, through which the problems induced by territorial fragmentation can be overcome. These mechanisms are best described by the term 'metropolitan governance' , i.e. various types of flexible, purpose-oriented co-operational arrangements in metropolitan areas, involving municipalities, different governmental agencies at various levels as well as private service providers. Because they are heterogeneous conglomerates of actors and agencies from various backgrounds and with various legal competencies, they allow to produce public policies in metropolitan areas in a way that is relatively independent from the territorial boundaries of local institutions. Such network-like mechanisms of 'governance' appear as the major thrust of metropolitan policy-making today, and are generally considered as an effective means to address metropolitan problems².

2)This argument is supported by research on metropolitan areas throughout Europe .

However, this new approach appears as a mainly managerial answer to metropolitan problem-solving. In most mechanisms of metropolitan governance, the rationale behind the involvement of various actors and institutions is an incremental mixture or criteria such as technical know-how, control of financial resources, as well as diffuse impressions of 'concernment'. The main idea is to 'get things done'. As such, the new metropolitan governance has raised uneasy feelings regarding democratic quality: with the major emphasis residing on implementation and hence on the 'output-side' of the policy process, the 'input side' of policy formulation - the main locus of democratic control - seems to be put at distance and raises concerns about the "democratic legitimacy" of such arrangements. In the modern state, both representation and participation are strongly linked to the territorial structuring of political institutions. If governance makes possible an efficient production of metropolitan policies independently from the territorial institutional structure, it is unclear how metropolitan democracy - still very much confined to territorially defined institutions - is affected by this. In the light of the well-known tension between efficiency and democracy, there is thus reason to think that metropolitan governance profoundly affects the relationship between policies and democratic politics in metropolitan areas , and several observers have suggested that a democratic deficit resides within current arrangements of metropolitan governance .

The objective of this paper is to address this question of the "democratic deficit" with respect to recent developments in metropolitan governance in Switzerland. We will start off with a brief presentation of the structural characteristics of the Swiss metropolitan areas and the issues of governance therein. Second, we we will introduce a conceptual framework in order to study issues of democracy with respect to metropolitan governance. On the one hand, this entails a specification of the somewhat fuzzy term of 'governance': more precisely, we suggest to think of government and governance in terms of an ideal-typical³ continuum. On the other hand, we will try to specify the issue of democracy in the context of governance by the formulation of four working hypotheses with which this issue can be empirically examined. In the third section of the paper, we will use these hypotheses to analyse some of the major mechanisms of policy-coordination in four metropolitan areas in Switzerland. In the conclusion, we will discuss the consistency and the usefulness of the used framework in the light of the empirical analysis.

3) In the sense of Max Weber's concepts of abstract ideal types that do not exist in reality but serve to classify empirical observations according to the resemblance to various ideal-types.

1. Metropolitan areas and governance in Switzerland

a) Metropolitan areas

As in most other industrialised countries across the world, urbanisation has profoundly transformed the socio-economic and the territorial structure of Switzerland . The growth of cities intensified in parallel to the increasing industrialisation in the second half of the nineteenth century. As elsewhere, socio-economic activities started to extended well beyond the traditional community boundaries, thereby giving rise to a new socio-spatial phenomenon: the urban agglomeration, characterised by a continuity of the built environment and with a strong orientation of socio-economic activities on a core area (usually: a traditional medieval city). The development of powerful systems of mass transportation since World War II gave rise to suburbanisation and periurbanisation, i.e. the exodus of wealthy families from the cities to the suburbs, which contributed to a further expansion of the urban space. In order to account for the ever extending urbanised space, the Swiss federal statistical office created the territorial category of metropolitan areas ('Metropolitanräume/Aires métropolitaines') at the occasion of the 1990 population census . These metropolitan areas are basically defined as groups of agglomerations linked to each other by a flow of persons and economic exchanges⁴ . They are exclusively statistical categories and have no institutional or political function. They account for the existence of a functionally integrated territory, that extends beyond communal, cantonal and even national borders. Five metropolitan areas were identified on the basis of the 1990 population census (see map in the appendix): Zurich (1.4 million residents, 172 communes located in six different cantons), Geneva-Lausanne (0.9 million residents, 150 communes located in two different cantons and two different French Départements), Basel (0.6 million residents, 72 communes located in four cantons, one French Département and one German Bundesland), Bern (0.5 million residents, 34 communes all of which located in the canton of Bern), Ticino (0.4 million residents, 119 communes located in the canton of Ticino and two Italian Provincie). As elsewhere, the existence of functionnally interdependent urban territories is thus a reality in modern Switzerland.

4) More precisely, agglomerations are defined according to a variety of criteria ranging from continuous built environment, over economic activities and centrality functions to commuter movements. Metropolitan areas are defined by the flow of commuters between agglomerations [Schuler, 1994 #439: 95-100; 187].

b) Issues of governance

In spite of this urban growth and the increasing connectedness of territories, the institutional structure of local government has virtually not changed since the beginning of the 20^th century. Only in a period between 1893 and 1934, some major cities (mainly Zurich and Geneva) have integrated suburban communes. Back then, suburban communes with feeble resources agreed to merge with prosperous cities that were lacking space for further development. But since the 1930s, such mergers have been very rare. Indeed, most suburban communes have consolidated their economic basis, whereas the core cities have gradually run into financial difficulties due to the intensification of the suburbanisation process (loss of wealthy taxpayers, and in parallel a concentration of socially disadvantaged segments of the population) . As other european countries, Switzerland has made a big step towards the scenario of «shrinking cities» , surrounded by a 'grease belt' (Speckgürtel) of rich suburbs. It is clear that in such a configuration, territorial reforms in urban areas are definitely not on the agenda: rich suburban communes ferociously oppose any step towards merging with financially distressed core cities . As a consquence, institutional fragmentation of metropolitan areas is high in Switzerland, and, in spite of the supposedly further ongoing urbanisation process, fragmentation is very likely to remain high in the years to come.

Research on local government in Switzerland consistently shows that supra-local policy-coordination in urban areas is mostly achieved through voluntary cooperation between communes. In their national survey on local government strategies - the first one to be conducted in Switzerland - Meylan et al. showed that intercommunal cooperation, i.e. joint implementation of certain public services, have a long tradition in Switzerland. Two decades later, Arn and Friederich confirm this observation, and even found evidence of an intensification of voluntary intercommunal cooperation in urban areas. Arn and Friedrich also show that a main argument put forward in favour of intercommunal cooperation is that it relies mainly on purpose-oriented arrangements which allows a variable territorial geometry according to the issue being addressed. Indeed, it is clear that the spatial dynamics and extension of policy issues can greaty vary and thus requires territorially flexible responses. More recently, a survey of communal secretaries by Ladner et al. shows that voluntary intercommunal cooperation has steeply increased in a period from 1993 to 1998, especially so urban areas .

Voluntary issue-oriented cooperation between communes, rather than institutional consolidation into some supra-local institutional body is thus the rule in urban areas in Switzerland. And, considering the local political elite's notorious opposition against amalgamations , voluntary intercommunal cooperation is likely to remain the main thrust of supra-local policy coordination in Switzerland.

Local government scholars in Switzerland have repeatedly emphasised issues of democracy with respect to such schemes of voluntary intercommunal cooperation. Research on local government in Switzerland consistently shows the absence of citizen involvement in structures of intercommunal cooperation. In their analysis of various legal forms of intercommunal cooperation, Arn and Friedrich have shown that the overwhelming majority of cooperation schemes only entail a representation of the executives of member communes, but no direct citizen participation. Similarly, the local elite generally displays a clear opposition against the introduction of citizens' rights in intercommunal cooperation. In the Meylan et al. survey, two thirds of the surveyed local executives were opposed to comprehensive democratic rights at an intercommunal level.

Thus, this brief overview suggest that the current trend of metropolitan policy-making in Switzerland is made of purpose-oriented voluntary coordination schemes where a strong position is confined to the executives of single municipalities. However, one counterpoint to this general picture must be mentioned: a law voted in 1994 in the canton of Fribourg, that lays the legal bases to create encompassing democratic institutions at the level of agglomerations. But - despite general recognition that this is a very avantgardist law - no such body has been created till now.

2. A conceptual framework to analyse the links between governance and democracy

a) How to distinguish government from governance

The widespread use of the vocable of governance in recent years by political and economic decision-makers, as well as by social scientists pertaining to various schools of thought and traditions of research has resulted in the emergence of many different understandings and definitions of this term⁵. Drawing on a major distinction made by Leresche , our understanding of the term of governance is not normative - such as the World Bank's good governance for example - but analytic. The main purpose of its use is to analyse interactional systems with respect to the political organisation of power, as well as with respect to public policy-making. In other words, our use of the concept of governance is to identify analytical dimension which allow us to detect whether or not, and to what extent, the empirical reality observed should be called governance and distinguished from other forms of regulation - and especially from traditional government which most tenants of the governance concept implicitly consider as being something like the antipode of governance. Hence, what are the specific features of governance?

5) For a review, see Le Galès , Rhodes , Stoker and, more recently, Leresche .

Going through a certain number of litterature on governance as well as through different definitions of it, seeking for a common core (despite the differences) of what governance is all about, should help us to define specific features of governance. Rhodes succinct definition of governance as 'self-organizing, interorganizational networks' provides a good starting point, in that it sets the tone of virtually all political science and sociological approaches to governance, namely that the policy process is envisioned as a 'bargaining game between different types of actors' (Dowding 1995: 147). Governance thus is about organising public action beyond the state and supervising, co-ordinating and steering a network of actors to produce certain outcomes. It is about bringing together all relevant actors in order to effectively and efficiently implement certain measures and services. The state tends to provide nothing more than "contextual steering" by creating incentives to exchange experiences and pool resources. State agencies may remain in charge of the network management by taking on the role of "public managers", which implies a certain absence of hierarchy. According to Stoker '[g]overnance involves working across boundaries within the public sector or between the public sector and private or voluntary sectors. It focuses attention on a set of actors that are drawn from but also beyond the formal institutions of government. A key concern is processes of networking and partnership.'

In sum, although using a varying terminology, these various discussions of the term of governance (implicitly) refer to a relatively stable grid of analytical elements serving to classify forms of interactions in the organisation of political power and policy-making including at least actors, conventions of interaction, distribution of power, structure and function. In a similar way Van Waarden specifies policy-networks through (1) actors, (2) functions, (3) structure, (4) degree of institutionalisation, (5) conventions of interaction, (6) distribution of power, and (7) strategies of public administration. Van Waarden uses these seven elements to characterise various types of policy networks and classifies them in different types of policy-networks, going from statism, to captured statism, clientelism, pressure pluralism, sectoral corporatism, macro corporatism, state corporatism, sponsored pluralism, parental relations, iron triangles and issue networks .

This characterisation is very interesting for our purpose, and we mainly draw on Van Waarden's contribution for defining specific features of "governance" (issue networks) as an antipode to "government" (statism)⁶ . However, with van Waarden's typology, it might be difficult to account for national or local context-effects (such as national traditions or sectoral policy styles) in distinguishing different forms of policy regulation. For instance, 'statism' (government) may well be an appropriate characterisation of policy-making in some countries (e.g. France) or some policy fields (e.g. national defence), whereas 'issue networks' (governance) have more tradition in other countries (e.g. Switzerland) or other policy fields (e.g. social policy). Thus, van Waarden's typology might be too rigid to account for transitions within national contexts that are already characterised by more or less negotiated policy-making, i.e. by a general "penchant" for either government or governance. In that respect we draw in addition on a further distinction proposed by Rhodes in which he suggests that types of policy regulations can be categorised along a continuum of interaction-types ranging from state-orchestrated to self-organising - hence: from government to governance. This puts the emphasis on changes and transitions within policy fields or territorial contexts. In other words, the question to be asked is not: is there government or governance in a given policy field or territory? but rather: does a given policy field or a given territory evolve towards government or towards governance? On the other hand, when it comes to the selection of case-studies in different policy-fields in order to study issues of democratic legitimacy, and theses selected case-studies should be characterised by more or less governance-like interactional systems, the idea of a continuum of governance does not bring us much further, because of the unlimited number of positions in between the extremes. Probably we should think of a typology that distinguishes 4 or 5 different positions on the continuum. For the moment we will keep this idea in mind and focus on the specification of definitional elements for the two extremes: governance and government.

6) In van Waarden's conceptualisation, 'statism' can be seen as an ideal-typical opposite to 'issue networks', in that they differ on almost all dimensions.

In synthesis, based on van Waarden's typology, we suggest the following set of features with which we can position empirically observed interactions on a continuum governance - government:

Table 1: Features setting apart governance from government ⁷

Government	Dimension	Governance
very limited number of participants mainly state agencies	Actors	high number of participants public and private actors
few consultation No co-operation in policy-formation / implementation policy issues broad	Function	more consultation Possible co-operation in policy-formation / implementation narrow policy issues
closed boundaries territorially defined boundaries involuntary membership	Structure	extremely open boundaries functionally defined boundaries voluntary membership
hierarchic authority, interlocking leadership Adversial interactions / conflictual relations informal contacts secrecy	Conventions of interaction	horizontal consultation, intermobility Consensus on technocratic norms / co-operative relations extremely informal contacts openness
high autonomy of state re society (steered organising) / state dominant no capture of state by societal interests no balance or symbiosis between actors	Distribution of power	low autonomy of state re society (self-organising) / diffuse domination of state diffuse capture of state by societal interests balance or symbiosis between actors

7) The table is structured according to van Waarden's analytical elements and shows those categories where 'statism' (government) show different characteristics from 'issue networks' (governance).

b) Governance and democratic legitimacy: four working hypotheses

The section above makes clear that 'Governance' differs from 'Government' in great many respects. The transition from government to governance thus corresponds to a significant change in the regulation of interaction between policy relevant actors. For this reason, when it comes to questions of democratic legitimacy of public policies - tightly linked to the functioning of the traditional democratic Rechtsstaat - scholars have ever viewed governance with a somewhat uneasy feeling. In western industrial democracies, the legitimacy of the state and state policies is generally seen to rest upon transparent procedures that give good reasons to believe that decisions made and policies implemented are, ultimately, the result of public deliberations among free and equal citizens . This is the raison d'être for mechanisms assuring democratic accountability and citizen control over state activities and public policies . From that perspective, it becomes clear that, when looking at the transitions from government to governance, one must keep an eye on eventual consequences that new mechanisms of governance may bear not only at the 'output-side' of policy implementation, but also at the 'input-side' of formulation and democratic control of public policies. This point is evidently not overlooked in the governance debate. As Rhodes, for example, points out, governance is likely to pose a considerable 'challenge for democratic accountability' .

In our view, in the ongoing intense debate on this point, two main lines of reasoning can be distinguished, which we propose to term the pessimistic and the optimistic view on governance and democracy. In the light of this debate we can derive four working hypothesis, that allow the assessment of issues of democratic legitimacy with respect to government, respectively governance-type regulations in different policy fields and different metropolitan areas.

The pessimistic view

One line of reasoning mainly rests upon the classic argument of democracy theory, according to which democratic political systems are structured by an intrinsic tension between authenticity and effectiveness of policies , i.e. quality of inputs versus quality of outputs. In the extent to which governance mechanisms are mainly concerned with 'making things happen', i.e. to increase the effectiveness of policies at the output-side, there is reason to think that this will happen at the expense of authenticity at the input side. Tenants of this line of reasoning thus argue that transitions from government to governance threaten democratic quality and legitimacy. More precisely, they have considered that mechanisms of governance could be a menace to two important fundaments of democracy: democratically elected bodies, as well as the political community itself.

First, they argue that governance relativises the weight of instances of democratically elected bodies in the policy process. Governance relies on self-governing networks associating various public and private actors and agencies, which, by the same token, leads to a dispersal of political power and to a loss of importance of traditional elected bodies in making decisions that are relevant to a policy field. More precisely, some observers have pointed out that the increasing participation of non-government organisations ('quangos') narrows the responsibilities exercised by formally elected authorities thereby modifying the relationships between voters and the government: lines of accountability become murky . Some claim that networks of governmental and non-governmental actors where the main criterion for membership is technical know-how, leads to a technocratic style of policy making where elected bodies increasingly loose grip . And, referring particularly to structures of multi-level governance, others have observed that networks associating governments, public administration and non-government from different state levels in a non-hierarchical way, contribute to a disenfranchisement of parliaments and councils at all levels .

Second, always according to the pessimistic view, it is argued that governance may affect the very fundamentals of community-identification. Networks of governance are mostly concerned with and structured by single policy issues, thus contributing to societal fragmentation and altering the ways through which the legitimate version of the common interest is identified and negotiated . As some researchers have noticed , governance conceives the citizen primarily as a taxpayer and a consumer of various public goods, whereas his/her role as a political subject seems somewhat in retreat. In their view, this fosters a conception of the political as just another sphere for the pursuit of individual goals, thereby calling into question the idea of the political community as a collective being with a public interest that is different from the sum of particular interests. In other words: governance would tend to make us forget that citizenship in the democratic state rests not only on the republican ideals of liberty and equality, but also fraternity .

To us, this line of reasoning seems basically a pessimistic perspective in the sense that its thread is the willingness to examine threats to democracy and legitimacy by mechanisms of governance and to ring alarm bells if necessary.

From these arguments we derive the following two hypothesis:

– H1 : Unlike government, governance leads to a relativisation of the role of democratically elected bodies

– H2: Unlike government, governance leads to a fragmented vision of the political community

The optimistic view

A second line of reasoning with respect to the implications of governance for democracy rolls up the issue from another end. Instead of centring around the tension between quality of inputs versus quality of outputs put forward by classic democracy theory, the second line of reasoning focuses on another key feature of democratic politics: the relationship between the state and civil-society, i.e. the Inexus of associations through which people organise independently to manage their own affairs and which can also be a channel of influence upon government and a check on it powers' . It emphasises the idea that popular control over public policies is not only exercised by elected bodies in a top-down way, but also directly by citizens and their associations. Tenants of this perspective insist on the prospects of governance with respect to direct citizen involvement and participation in public policies, independently from representative elected bodies. Two main elements have been examined with respect to how transitions from government to governance affect the relationship between civil society and the state: the pluralisation of policy-making by the pluralisation of policy-networks, the emergence of new instances of deliberation and dialogue between state and citizens.

First, drawing on the tenet of 'associative democracy' , it is argued that self-governing networks, that include not only state agencies but also varieties of non-government organisations and associations, are an important vector of pluralism and civic culture. The presence of associations in these networks increases the influence of the civil society on public policies and fosters civic culture. The involvement of non-government organisations and associations in the enforcement and administration of policies would result in increasing the points of view and the interests served by a particular public policy. Thus, recruiting the energies of citizen's organisations into public governance could be a step away from an oppressive state and towards a more egalitarian-democratic order . In that sense, through the link of association involvement in public policies, transitions from government to governance have been viewed as an opportunity for empowering the citizen and pluralising the state .

Second, it is argued that governance structures may foster discursive processes and thus increase deliberative quality of public policies. The reason for this lies in the augmentation of interdependencies due to the great number and variety of actors and agencies involved in governance structures. Hierarchies are mostly absent, and when they are not, veto power of single actors and agencies are often considerable. Hence, negotiation and compromise, or deliberation and consensus appear as the only way of achieving agreements on the course of collective action in self-governing networks. In this sense, the emergence of new spheres of negotiation and deliberation, such as neighbourhood forums, citizen juries, user boards etc. has been interpreted as an increased importance of discursive processes in the formulation and administration of public policies as consequences of the transition from government to governance .

The central tenet of this second line of reasoning is thus to emphasise the potential for democratic renewal "from below" which resides in the new mechanisms of governance. This view can be termed optimistic, in the sense that it set out to examine whether the transition from government to governance opens up the state towards civil society and, if so, where these new openings occur and what their potential is for reinvigorating democratic civil society.

In the light of these arguments, we are left with two optimist hypothesis:

– H3: Unlike government, governance leads to pluralism and to the empowerment of citizens.

– H4: Unlike government, governance fosters deliberation and consensus as a main mode of decision making.

3. Metropolitan governance and democratic legitimacy: a primary test of the framework with the empirical material collected so far

In our currently onoing research project, we set out to discuss the above presented hyoptheses through the study of four specific policy fields where mechanisms of metropolitan governance are common: provision of medical and social services for drug users, public transport, cultural institutions and water provision. These policy-fields were selected according to two lines of criteria: on the one hand very politicised policy-fields (drug-policy, public transportation) versus policy fields with very few debate (water provision and cultural institutions); and on the other public services that are provided in a reticular way (public transportation, water provision) versus services provided punctionally (cultural institutions, medical and social services for drug users). As the data collection is still ongoing we will dedicate the remainder of the paper to report a primary test of these hypotheses on some major schemes of territorial policy-coordination in four Swiss metropolitan areas⁸. In this sense, the idea of this analysis mainly is to submit the analytical framework to discussion. We will first use the analytical categories used to distinguish governance from government type regulations (cf. government:

Table 1 ) to try to position these various mechanisms on the continuum government versus governance. In a second step, we will test each of the four working hypotheses by applying them to the various mechanisms.

Four metropolitan areas in Switzerland are under scrutiny in this research: Zurich, Lausanne-Geneva, Bern and Ticino⁹.In all four metropolitan areas , there exist a myriad of institutions that aim at the coordination of policies with a metropolitan scope, i.e. that go beyond the boundaries of single territorial entities (communes or cantons) that compose the metropolitan areas. For the purpose of this paper, we will limit ourselves to the examination of those coordination schemes with a logic going further than sectorial preoccupations (related to one specific topic). According to their territorial scope, two types of such schemes can be identified.

8) The analysis rests on a press articles, information material produced by the various stakeholders, as well as on several informative interviews conducted with the secretaries of the mechanisms under scrutiny.

9) The case of Basel was left aside because of its extreme institutional complexity which would have required important supplementary means of research. Moreover, because of this institutional complexity (especially the close collaboration schemes between the two half-cantons of Basel-Stadt and Baselland) it is doubtful whether the Basel case can really be considered significative for the discussion of metropolitan governance issues in Switzerland.

First, intercommunal schemes - associating the communes located in the metropolitan area - exist in every of the four areas under scrutiny and they are all private associations according art. 60 ff. Swiss civil law: In the Geneva-Lausanne metropolitan area, there is the Association des communes genevoises (ACG) associating the totality of the 45 communes located in the canton of Geneva, as well as the Communauté de la région lausannoise (COREL) reuniting the communes located in the agglomeration of Lausanne. In the Bern metropolitan area, there is the Verein Region Bern (VRB), linking the communes located around the city of Bern. The Gemeindepräsidentenverband (GPV) was found in the Zurich metropolitan area, and the Associazione dei communi urbani (ACU) reuniting urban communes of the canton of Ticino.

Second, except for the Zurich case, there exist higher level coordination schemes which associate various types of private and public actors with various territorial scopes (communal, cantonal, even international). For the Bern region, there is the Espace Mittelland, an association of five¹⁰ cantons reunited to defend the interest of the greater Bern area located in between the two very strong urban centres of Zurich and Geneva-Lausanne. In the Geneva-Lausanne metropolitan area, appeared first the Commission mixte franco-genevoise put up in 1973 by the Swiss and French government with the scope to coordinate transnational relations in the Geneva area, it involves the city and the canton of Geneva, as well as the French Departments of Ain and Haute-Savoie. In 1987 was also created the Conseil du LEman involving public actors located on the Swiss and French territory around the lake of Geneva (same actors as in the Commission mixte franco-genevoise plus the cantons of Vaud and Valais). Last, but not least in the Ticino area, there is the Regio Insubrica, a transnational organisation involving public and private actors from Ticino and the three adjacent Italian Provinces (Como, Varese, Verbano-Cusio-Ossola).

10) Bern, Fribourg, Neuchâtel, Solothurn, Jura

Table 4 and 5 in the appendix show various elements of information on these coordination schemes according to the above defined features, with which we suggest to distinguish government- from governance-type regulations. It is clear from these tables that none of these metropolitan coordination schemes can be considered as purely 'government'-like in the traditional weberian sense of constitutionally structured political institutions exercising a power monopoly on a given territory. In fact, all of them are private associations from a legal point of view and include quite a high number of participants who are themselves state agencies.

However, the coordination schemes do differ in some of the central categories used to distinguish government from governance. In general, all intercommunal associations (i.e. ACG, COREL, VRB, GPV, ACU) include only public actors and membership is (except for the ACG and the GPV) volontary. They integrate generally a high number of actors, but private actors are associated only as experts, without having the status of member and the right to vote. This is also the case for the Espace Mittelland and the Conseil du Léman. The Commission mixte franco-genevoise instead has one private actor member (the groupment des frontaliers de l'Ain et de la Haute-Savoie) . Last but not least, the Regio Insubrica definitely appears as the coordination scheme which is most 'governance-like', with its extremely heterogeneous types of members (public, private, associational), its extremely open boundaries and its explicit attempt to find a balance between all the actors, independently from whether they are public, private, local, supra-local, Swiss or Italian. The boundaries of these institutions are generally open (except for the ACG and the GPV), in the sense that they encourage the adherence from supplementary territorial entities. On what their function and political role is concerned, the intercommunal schemes of coordination have a lot of similarities: they all have a consultative role of policy-proposition. The Association des communes genevoises also has a role of policy implementation, it manages the computer equipment for the communes, organises para-school activities for children and burns animal carcasses for the communes. The intercantonal (-national) institutions also have a role of consultation, but whereas the Commission mixte franco-genevoise has a relatively significant political weight, the others are politically not very important. The conventions of interaction are in general very formal (except for the Gemeindepräsidentenverband) and put up in a way to balance the influence of the different territorial members.

What appears from this general overview, is that there are in fact few differences between theses mechanisms of coordination. The main difference resides in the weight of state-actors within these institutions. If we now try to classify these various coordination schemes on the continuum government versus governance, the most government-like would be the Association des communes genevoises, the Gemeindepräsidentenverband and the Commission mixte franco-genevoise, and the most governance-like the Regio Insubrica¹¹ .

11) It is evidently questionable to study these coordination schemes in general, but that one has to consider their role and functioning in concrete policy-cases, and study in detail the different features identified above, in order to specify whether the regulation in a given policy-case was more government- or more governance-like. This will be done in the remainder of the research, where we concentrate on the role of these coordination schemes in the four specific policy fields.

If we now want to discuss now eventual consequences of governance on democratic legitimacy, we need to test the four working hypotheses on these various contrasted cases. For instance for this paper we will compare four of the above mentioned coordination schemes. On the one hand, we will take a look at three government-like type of coordination schemes: the ACG and the GPV as representing the schemes with an intercommunal scope, as well as on the Commission mixte franco-genevoise as a government-type scheme with a higher level of territorial scope. On the other hand, we will examine the Regio Insubrica as a 'governance-type' contrast.

According to hypothesis one (Unlike government, governance leads to a relativisation of the role of democratically elected bodies) we can note that in the ACG and the GPV, communes are represented by elected officials (the mayor or an other member of the executive). The Commission mixte franco-genevoise involves representatives from the different Swiss and French territorial levels¹²  it includes for the French part even the presidents of the Parliaments of Ain and Haute-Savoie. There is thus a certain link of accountability from the electorate of the member institutions to the representatives in the associations. Furthermore, all important decisions taken by these associations must be ratified by the members. In practice, these coordination schemes therefore have almost no discretion, and the importance of local parliaments (who ratify decisions on the allocation of a credit, for instance) remains important. In the Regio Insubrica, the same is true for the public members who are represented by elected officials (the mayor in the case of communes, members of the provincial or the cantonal executive in the case of the canton and the Provincie). However, representatives of the private members (businesses, associations, etc.) are not designated according to any transparent procedure: the decision whom to send is totally left to the member. Concerning the position in the policy process, the Regio Insubrica explicitly states that it does not want to be a substitute to the elected bodies. The Regio only has a very small budget - covering costs for its secretariat - and needs to search for special funding from its members, when it wants to implement activities for which some budget is needed. Hence, where the transfer of public resources to the Regio is concerned, the role of local governments and parliaments remains important.

12) For the French part: The préfets de la Rhône-Alpes, des Départements de l'Ain et de la Haute-Savoie, the Consul general de France in Genenva, as well as the presidents of the Conseil généraux de l'Ain et de la Haute-Savoie / For the Swiss part: Members of the executive of the commune and the canton of Geneva, a high functionary of the Federal Department of External Affairs, and a representative of the Association des communes genevoises.

Hypothesis two (Unlike government, governance leads to a fragmented vision of the political community) would require an inquiry among the members of the political community (the citizens) to see whether fragmented or unitarian vision prevails - and it would be even more difficult to assess whether this is the result of the activity of any coordination scheme... However, we can assess the vision of community that is portended by the participants of these coordination scheme. From our interviews, it appears that in the case of the ACG and the GPV, conflicts between representatives of various types of communes (e.g. those receiving money in the cantonal compensation mechanism versus those who contribute to the compensation mechanism) is quite frequent. In other words, one can not really speak of there being a vision of a new territorial identity that would be present in the discussions within the ACG. On the contrary, the rationale rather is one of a congregation of local interests, where common positions are rather hazardous. The case of the Commission mixte franco-genevoise seems intermediate: it has been put up with the very clear scope to coordinate transnational relations, so there is a concrete common interest (concerning for example public transportation) and a vision of community.

A totally different picture is given by the Regio Insubrica. Indeed, decisions and debates seem rather consensual, grounded on the basic consensus that there is a vague common interest to jointly developing the Region against the concurrence of Milan. However, the fact that there have been no important conflicts at the Regio Insubrica may also result from it concentrating mainly on topics of promotion, whereas distributive issues havenIt been tackled so far.

According to hypothesis three (Unlike government, governance leads to pluralism and to the empowerment of citizens) and according to the government-type coordination schemes ACG and GPV, we can note that it is not plausible to speak of them as a vector of pluralism or for empowerment of the citizens. Indeed, it appears as a coordination scheme of representatives of the communal executives: neither are there any civil-society associations represented that could transmit societal pluralism into the coordination schemes, nor are there any new procedures of citizen consultation or involvement. As for the Regio Insubrica, there is involvement of civil society associations: besides public institutions and private business, members include about a dozen civil society associations with full membership rights. However, formal citizen empowerment or involvement in the Regio's activities has not taken place so far, either.

As to what hyopthesis four (Unlike government, governance fosters deliberation and consensus as a main mode of decision making) is concerned, we can note a striking similarity across all coordination schemes according to the mode of decision making (majoritarian versus consensual). From a formal point of view, the statutes of each coordination scheme allow majority decisions in the various organs. However, according to our interviews, majoritarian decisions are extremely rare. Decisions are taken at unanimity, or they are not taken at all: when participants cannot reach a consensus on some topic, they tend to drop it. This general mode of functioning does not differ between the three government-type of coordinations schemes and the Regio Insubrica as a more governance-type scheme. This can be explained by the fact that the competences of these various coordination schemes are very limited; no member can be forced to adopt any behaviour against her/his will. In such a context, it does indeed not make much sense to take majoritarian decision who are likely to be followed only by those who backed them...

It is clear that this analysis of the functioning of metropolitan coordination schemes needs to be deepened and refined if we want to make sound statements on the issue of democratic legitimacy with respect to metropolitan governance. In particular, it can be argued that the territorial coordination schemes examined here are "light-weights" and thus do not represent a good case for inquiry. Indeed, be they more government- or more governance-like: all of the coordination schemes examined here appear basically as mechanisms of consultation rather than of policy-making. It is thus questionable, whether it is appropriate to study supposed consequences of governance on democratic legitimacy on the basis of these coordination schemes which do not quite appear as "the real thing". In this sense, the remainder of our research project will be dedicated to the study of metropolitan governance specific to four policy fields where territorial coordination is highly relevant: medical and social service provision for drug users, public transport, urban cultural institutions and water provision. Nevertheless, even with the limited analysis on the metropolitan coordination schemes presented above, it is possible to draw some preliminary conclusions on the issue of governance and democratic legitimacy, as well as on the analytical framework used to examine this issue.

4. Conclusion: the issue of democratic legitimacy and the usefulness of the analytical framework

Concerning the issue of democratic legitimacy with respect to metropolitan governance, the above analysis yields a picture that somewhat confirms the "optimist's view". One pessimist working hypothesis is clearly falsified (H1): we did not find that the role of democratically elected bodies was any less important in the context of the governance-type coordination scheme than in the context of the government-type schemes. Thus, one major point of the pessimist's view does not correspond to the empirical reality. Furthermore, one optimist working hypothesis (H3) tends to be rather confirmed: the governance-type coordination scheme with its heterogeneous membership structure is clearly more open to civil-society. Hence, one major optimist point reveals quite true with respect to the empirical situation.

The results on the remaining two hypotheses are less clear. With respect to H2, according to which governance puts forward a fragmented vision of the political community, the analysis shows that both in the government- and the governance-type schemes, there is a certain identification of participants with a territory that goes beyond the scope of the local corporations to which they initially belong. However, because of there being more often conflictual situations among participants, it may well appear that this collective identification is more challenged in the government-type schemes. Thus, even if H3 also seems falsified at first sight in the extent to which collective identification definitely is not confined to government-type schemes, it may well be that their 'performance' in this respect is higher since situations are more often conflictual. Concerning H4, according to which governance-type regulations foster deliberation, the results show that type of decision making - formal and real - do not differ greatly from government- and governance-type of coordination schemes. However, the analysis also shows that citizen empowerment does not necessarily accompany governance-type of regulations.

In sum, concerning the metropolitan coordination schemes under scrutiny here, the optimist's view on governance is more corroborated than the pessimist's view.

Assessing the analytical framework

Throughout this paper, an analysis of the empirical evidence through the theoretical framework based on four working hypotheses on governance and democratic legitimacy appeared possible and feasible. In particular, although drawn from two lines of argument that we felt to be normatively oriented - pessimist or optimist - the confrontation with empirical evidence produced quite differentiated results. We thus conclude that these four working hypotheses can be fruitful for structuring and orienting empirical research on issues of democracy with respect to (metropolitan) governance.

Our analysis also suggests that it is far from evident that transitions from government to governance have intrinsic consequences on processes and procedures related to democratic legitimacy - such as it is suggested by major scholars of the field. In this sense, the argument of a supposed democratic deficit with respect to metropolitan governance is, at the minimum, simplistic and, at the maximum misleading. The empirical evidence shows that mechanisms of metropolitan governance can also be a vector for increasing the participation of civil society in metropolitan policy making. Thus, rather than speculating on a pessimist or an optimist outcome with respect to new mechanisms of metropolitan governance, the consequences of legitimacy with respect to governance need to be assessed through systematic empirical assessment. This is, undoubtedly, a first step in assessing the circumstances and contextual elements which condition that in some empirical situations one or the other line of argument reveals true (pessimist or optimist). We feel this to be the major research agenda related to the topic of governance and democratic legitimacy.

Appendix

Map 1 – Metropolitan areas in Switzerland

According to the definition of the Federal Office of Statistics (Schuler 1994) on the basis of the 1990 population census five metropolitan areas in Switzerland were identified: Zurich (1.42 million residents, 10 agglomerations, 172 communes located in six different cantons), Geneva-Lausanne (0.96 million residents, 4 agglomerations, 150 communes located in two different cantons and two different French Départements), Basel (0.6 million residents, 2 agglomerations, 72 communes located in four cantons, one French Département and one German Bundesland), Bern (0.4 million residents, 4 agglomerations, 34 communes all of which located in the canton of Bern), Ticino (0.4 million residents, 4 agglomerations, 119 communes located in the canton of Ticino and two italian Provincie).

Table 4 - Intercommunal schemes of policy-coordination

Name		Communauté de la région lausannoise (COREL)	Verein Region Bern (VRB)	Association des communes genevoises (ACG)	Associazione dei communi urbani (ACU)	Gemeindepräsidenten-verband (GPV)
Legal form		Private association (art. 60 Swiss civil law)	Private association	Private association	Private association	Private association
Actors	type	Only state actors	Only state actors	Only state actors	Mayors of urban communes of the canton of Ticino	Mayors of the communes of the canton of Zurich
	number	high	High	Very high	Medium	High
Function	Consultation	Yes	Yes	Yes	Yes	Yes
	Co-operation in policy formulation / implementation	In policy formulation	In policy formulation	In policy formulation and implementation	Only in policy formulation	Mainly in policy formulation, exceptionally in implementation (drug policy,
	Policy issues	broad	Broad	broad	Broad (no restriction to any specific issue)	Broad (no restriction to any specific issue)
Structure	boundaries	open	Open	Closed	Relatively open	Closed
	membership	voluntary	Voluntary	compulsory	Voluntary	Compulsory
	Frequency of interaction	high	High	medium	Low	High
Conven-tions of interaction	Authority vs consultation	consultation	Consultation	consultation	Consultation	Consultation
	Adversial/conflictual / vs Cooperative/ consensus on technocratic norms	conflictual	Cooperative	cooperative	Most of the time cooperative; conflictual on some issues	Most of the time cooperative and consensual; conflictual on some issues
	Formality of contacts	Very formal	Formal	formal	Formal	Formal
	Secrecy vs openness	openness	Openness	openness	Relative openness	Secrecy (explicitly refrained from public relation strategy)
Distribu-tion of power	Degree of autonomy / domination of the state	State dominant	State dominant	State dominant	State actors dominant	State actors dominant
	Capture of state by societal interest	no	No	no	No	No
	Degree of balance between actors	Balance between actors	Balance between actors	Balance between actors	Balanced (one member one vote)	Balanced

Table 5 - Intercantonal (-national) schemes of policy-coordination

		Conseil du Léman	Commission mixte franco-genevoise	Espace Mittelland	Regio Insubrica
Legal form		Private association (according to 1980 Madrid Convention on transborder cooperation)	Agreement between the French and Swiss government	Private association	Private association (according to 1980 Madrid Convention on transborder cooperation)
Actors	Type	Only state actors	Mainly state actors, 1 private actor	Only state actors	State and private actors (canton, province, communes, private firms, unions, associations)
	number	High number of participants	Limited number of participants	High number of participants	Very high number
Function	Consultation	yes	Yes	yes	Yes
	Co-operation in policy formulation / implementation	In policy formulation	In policy implementation and formulation	In policy formulation	In policy formulation
	Broadness of policy issues	Very broad (no restriction to any specific issue)	Relatively narrow (transborder cooperation)	broad	Relatively narrow (basically transborder cooperation, area promotion)
Structure	boundaries	open	Closed	Extremely open	Extremely open
	membership	volontary	volontary	volontary	Volontary
	Frequency of interaction	Extremely low	medium	low	Relatively low
Conventions of interaction	Authority vs. consultation	Horizontal consultation	consultation	consultation	consultation
	Adversial/conflictual / vs Cooperative/consensus on technocratic norms	cooperative	Consensus on technocratic norms, cooperative norms	cooperative	Cooperative, based on consensus (economic promotion and development of the area)
	(in) formal contacts	Formal contacts	Formal contacts	Formal contacts	Formal contacts
	Secrecy vs openness	openness	openness	openness	opennes (explicit PR strategy)
Distribu-tion of power	Degree of autonomy / domination of the state	State dominant	State dominant	State dominant	State not dominant
	Capture of state by societal interest	no	no	no	No (too many participants that one could dominate)
	Degree of balance between actors	Balance between actors	Balance between actors	Balance between actors	Balance (one member, one vote)