Planning for Interdependence. Linking Amsterdam, its Metropolitan Region and Europe

Ernest R. Alexander

1. Linking Interdependent Communities
2. Amsterdam in the Red Queen's Country

The two cases below have some things in common. One is their geographical focus: both take off from Amsterdam, and both put Amsterdam into a very specific context. But each follows a somewhat different trajectory. The first takes the Amsterdam metropolitan region as its point of departure, moving at once inward (to the core city) and outward to encompass the metropolis' geographic and institutional environment: the Province of Noord-Holland and the Netherlands with its national ministries and agencies. The second focuses on the community of Amsterdam, but ignores its metropolitan-regional context and instead explores its links with the European Union (EU).

They also have a common research approach. Both are relatively detailed studies of institutional-organisational systems and their environments, the planning, policymaking and decision-making processes enacted there, and selected aspects of their consequences. Their level of analysis is somewhere between the individualised richness of ethnographic cases, and the abstraction of sociological studies: the relevant unit of observation is the organisation. Both apply what is discovered to learn something about the behaviour of the social units involved in the case, but both go on to make generalisations about their universes: about the role of consultation in successful planning and policy implementation (in the first case), and the importance of networking in European integration (in the second). These generalisations from specific case observations are scientifically and logically untenable, of course, but they may be thought-provoking.

Finally, they have a common theoretical frame-of-reference: both explore what I have called co-ordinative planning. This is planning that is in essence anticipative co-ordination between organisations, as distinct from planning viewed as rational decision-making by individuals or as communicative interaction between them, or planning seen as community-wide discourse and conceptual frame-setting (Alexander 1998). Though all planning is associated with organisation (Alexander 1992), this is especially the case for co-ordinative planning, where organisations are the relevant actors, and which can be understood as one form of interorganisational co-ordination (IOC).

According to the 'Structuration theory of IOC' (Alexander 1995: 66-75), it is actual and perceived interdependence between organisations which makes them act together, which makes co-ordinative planning necessary and elicits the IOC structures in which co-ordinative planning is institutionalised. These interdependencies, then, implied or explicit, and the IOC strategies and structures that make up co-ordinative planning, are the common subject of the two case studies presented here. In the first case, the stimulus for metropolitan-regional planning is the functional interdependence between the communities in the Amsterdam urban region, which has been institutionalised in various forms of IOC structures, ranging over time from consultative ones to attempts at regional governance. In the second, the growing sense of the community's role in its broader European context has stimulated a rich array of networking activities, to link Amsterdam to the EU in ways that were unthought of a decade before.

Ernest R. Alexander

10 March 1997

1. Linking Interdependent Communities

Metropolitan-Regional Planning in Amsterdam

Introduction

Metropolitan areas are recognised as a special kind of region, for analytic (e.g. statistical), administrative and planning purposes. Metropolitan-regional planning goes on all the time, in hundreds of places all over the world, from Albuquerque to Zagreb. Yet it has not received much attention, especially in making the lessons of experience available to current and prospective practitioners.

When planning metropolitan regions is discussed, it is often as an adjunct to metropolitan-regional governance. This is the dominant perspective of the (rare) prescriptive approaches, such as the one in the American Planning Association's 'Green Book' (McDowall 1986), and of the many cases like the descriptions of Chicago, Edmonton and Toronto area planning included in a comparison of US and Canadian metropolitan areas (Hemmens and McBride 1993; Thomas 1993; Frisken 1993). Or planning is subsumed under metropolitan-regional development policy, as in case studies of Amsterdam, Birmingham, Boston, Frankfurt, and Rotterdam (Harding 1994; Evans 1994; Gelfand 1993; Kunzmann and Land 1994). While they provide much useful information, such case-studies fail to present the experience of metropolitan-regional planning in its full complexity.

Complexity, however, is the salient characteristic of metropolitan-regional planning wherever it is done, and it is perhaps the aspect that is most important to understanding the planning process, and the reasons for its successes or failures. This is implied if we view metropolitan-regional planning as one form of interorganisational co-ordination (IOC). While planning can, indeed, be defined as anticipatory co-ordination (Alexander 1993: 328-9), and much planning is concerned with IOC (Alexander 1994: 343-8), metropolitan-regional planning is distinctive.

The major task of metropolitan planning is to co-ordinate the policies, decisions, and actions of the governments, agencies, organisations, firms and households which determine the development of the metropolitan region and form its physical and socio-economic environments. This makes metropolitan-regional planning an exercise in IOC par-excellence, suggesting the IOC perspective (Alexander 1993) as a useful way to analyse metropolitan-regional planning, evaluate its performance, and understand its results.

It is the interdependence of the local communities and their populations in the area that makes planning the metropolitan region necessary. The way in which prevailing consciousness of this interdependence, and perceptions of its consequences, are institutionalised, is reflected in the IOC structures and strategies employed in metropolitan-regional planning. The case of the Amsterdam region may test the proposition that a critical factor in the success or failure of metropolitan regional planning is the way planning is institutionalised in IOC structures, how these perform and 'fit' into their environments.

Background

Amsterdam and its hinterland cover much of the north wing of the Randstad -the crescent-shaped conurbation extending across the Netherlands, from Rotterdam at one end to Utrecht at the other, which, with its population of around seven million, ranks with London and Paris as a major European metropolitan agglomeration. The Amsterdam metropolitan region extends over nearly 117,000 hectares and its 1995 population of 1,386 million is expected to grow to nearly 1.7 million by 2015. It contains sixteen communities, ranging from the city of Amsterdam, through large suburban municipalities with populations from 50- to nearly 100,000, to still almost rural villages like Zeevang with its 4,236 people.

In spatial terms, this region is still focused on Amsterdam as its traditional core, with clear traces of its historical growth from the fishing port behind the 'Dam' on the Amstel river which was chartered by Count Floris of Holland in 1275. The city's medieval centre and its seventeenth and nineteenth century expansions are still much in evidence, while the region grew with the city with urbanising satellite communities that often also began as fishing villages on the inland waterways. Reclamation of land from the sea increased its area multifold over the centuries, creating the 'polder' landscape so typical of North Holland. The most recent major project was the drying of the Haarlemmermeer, now the site of Schiphol international airport. But today, though there are still rural parts like the Waterland polders and the coastal dunelands, they are residual in what is a highly urbanised metropolis, with its localities tightly linked to its node with well developed road and rail networks.

The region's economic base, too, is typical of an advanced post-industrial society. It has a role in the global economy as a major gateway to Europe via its Schiphol international airport. Typically, too, this metropolis experienced declining industry and growing tertiary and quartenary sectors which today provide nearly three-quarters of its 600,000 workplaces. In these sectors financial and related services lead, with strong distribution and tourism, a flourishing information-knowledge 'industry', and diversified retail and commercial services. Still, restructuring the metropolitan region's economy is not problem-free, and facilitating its economic development is an important planning objective (DRO 1994a; ROA 1995).

Besides the usual system-maintenance tasks, the metropolitan region's policy makers and planners confront difficult decisions. Many of these involve the distribution of expected urban growth, and trade-offs between economic development, environmental and quality-of-life considerations. The expansion of Schiphol airport (discussed below) is an example. Planning here takes place in a social and governmental context that has some characteristics which are perhaps unique. The Dutch, 'materialistic, conservative…(with their) virtues of reason, co-ordination…orderliness and consensus' seem almost tailor-made for planning (Schön 1997). The Netherlands are a Western European social democracy with a strong public sector and a well developed welfare state (from which, however, there has recently been some retreat).

Though there is a highly institutionalised system of planning in a unitary state, Dutch planning and governance are not hierarchical. The Netherlands has three levels of government-the national government, twelve Provinces, and local municipalities-with powers distributed between them in a form of 'co-government' in a 'decentralised unitary state'. The statutory planning system of physical-environmental planning ('spatial order' in Dutch), and sectoral planning as well, is divided between these levels, with the local land use plan (bestemmingsplan) being legally binding. The Dutch experience as a relatively consensual society is reflected in their 'politics of accommodation' and 'interwoven planning' (Faludi and Van der Valk 1994: 33-6, 38-40), which are well displayed in Amsterdam's metropolitan-regional planning described below.

Governance and Policy

As far back as the mid-nineteenth century, there was frequent friction between Amsterdam and its adjacent communities. Amsterdam perceived its neighbours as exploiters of a 'parasitic' relationship with it. They, in turn, while admitting their functional dependence, hated Amsterdam's metropolitan arrogance and feared its political domination. Their fears were not unfounded: many conflicts were resolved by annexations, which were widespread. The Province played only a weak role in mediating these differences, for which the Provincial statutory plan (the streekplan), limited to only indicative status, was ineffective. An example is the clash over development of the Bijlmer in the 1950s as an extraterritorial expansion area for Amsterdam social housing, which was finally resolved by annexing the area to Amsterdam (Van der Valk and Wallagh 1994; Van der Veer 1996: 7-9).

In 1957 the Province of North Holland created a co-ordinating body called the 'Urban periphery Commission' (Stadsrandcommissie), including a provincial delegate as a mediator, and representatives of Amsterdam and its neighbour communities. This body prepared a structure plan for the metropolitan region, but the plan was ineffective as it needed member communities' agreement to incorporate it into their statutory plans.

These problems affect all the Netherlands' large cities, and the issue of metropolitan-regional governance has long been on the national political agenda. The 1950 Law of Community Regulations (Wet Gemeenschappelijke Regelingen - WGR), which was the framework for any possible structural solutions, was inadequate. It enabled intermunicipal co-operation in limited areas such as solid waste disposal or electricity supply, but offered no solution to more complex intersectoral metropolitan co-ordination problems.

The Amsterdam conurbation went beyond this law, when the Informal Agglomeration Consultation (IAO - Informeel Agglomeratieoverleg) was established in 1972 as an arena to address regional issues. The IAO included 25 local governments, represented by their mayors or a councillor, and its agenda was often determined by proposals submitted by Amsterdam, on which the other communities and the Province made their comments. For some sectors, such as economic issues and housing, standing subcommittees were established, made up of aldermen holding the relevant portfolio in their communities. The IAO was basically an interorganisational group (Alexander 1995: 121-42) created to co-ordinate municipalities' policies on matters of common concern through information exchange and consultation. While the IAO served this purpose quite well, Amsterdam (and the national government) looked for more institutionalised structures to effect more intensive intermunicipal co-ordination (Van der Valk and Wallagh 1994: 11-23).

Through the 1960s and 70s there were other attempts at creating two-tiered regional governance for metropolitan conurbations (stadsgewesten), but the only one implemented was the Rijnmond Regional Council in the Rotterdam area. Meanwhile, the national government, eager to offer a formal framework for such co-operative efforts, introduced significant amendments into the WGR in 1985. These enabled the transformation of the IAO into its successor organisation, the Amsterdam Regional Consultation (ROA: Regionaal Overleg Amsterdam) in 1986. The ROA, designed to be a co-ordinating unit (Alexander 1995: 153-77) was governed by an interorganisational group: the mayors of the participating communities, chaired by the mayor of Amsterdam, and had a small professional staff.

In 1988 the national legislature appointed the 'Montijn Commission'to address continuing problems of metropolitan governance, but its solution was not regarded as politically feasible. The Commission rejected any options involving a new 'fourth level' of government (i.e. between the Provincial and the municipal levels), and looked for a solution at the local level. But its proposed 'agglomeration municipalities' around each of the four large cities required massive annexations which would be fiercely contested by the (now powerful) suburban municipalities.

The national administration, in consultation with the large city governments, continued to grapple with the issue of a feasible framework for metropolitan-regional governance. The national interest was not limited to the benign motive of improving metropolitan intergovernmental co-ordination. The government wanted to reverse the large cities' growing dependence on central government funding (Van der Valk 1994: 28-9; Faludi and Van der Valk 1994: 227-9), the result of increasing fiscal inequality between central cities and their suburbs (Van der Veer 1994).

Several Interior Ministry reports (1991 to 1993) articulated national government policy. Like the Montijn Commission, it wanted to avoid creating a fourth layer of government. Unlike the Commission, it looked for solutions at the Provincial, rather than the local level, and proposed the formation of 'City-Provinces' in the seven metropolitan areas (Van der Veer 1996: 3-4). At the same time, a form of metropolitan regional co-ordination was coming into existence incrementally in Amsterdam at the local level.

This was the result of agreement between the Provincial administration of North Holland and the city of Amsterdam on enhancement of the city's policy-making powers and reduction of Provincial supervision. Formalised in 1987, the intergovernmental contract required administrative decentralisation at two levels. At the provincial level, the Province delegated its statutory planning functions (with respect to Amsterdam's territory) to the Amsterdam municipal administration. This gave Amsterdam's structure plan the legal status of a provincial streekplan, with the Amsterdam Planning Commission taking over the plan-review functions of the Provincial Planning Commission.

At the local level, the compact provides for decentralisation of many of the city's administrative functions to sixteen neighbourhood sub-communities (deelgemeenten). These include planning functions such as preparation and updating of the local statutory land use plan (bestemmingsplan) and administration of local zoning and development control. In effect, this agreement involves a devolution of land use regulatory authority one level down: from the Province to the municipal administration (which now fills the Province's former role) and from the municipality to the deelgemeenten, which become, as it were, miniature cities (De Rooij 1994).

Implementation of this contract (in progress at the time of this writing) also set the stage for regionalisation in the Amsterdam area, as envisaged by national government policy, by breaking up Amsterdam into separate communities. Limiting the city's power and reducing its dominance was a precondition for the other communities' agreement to participate in any kind of formal regional governance (Faludi and Van der Valk 1994: 192-4).

A Framework Law (Kaderwet bestuur in verandering) was enacted in 1993 which allowed metropolitan areas to establish directly elected regional governments. These would take over what were previously Provincial responsibilities in regional structure planning, housing allocation, regional transportation planning and policy, and economic and environmental policy (Van der Valk 1994: 34-5; Van der Veer 1996: 3-5). At the end of a transition process (scheduled for 1/1/1998), metropolitan areas like Amsterdam's would become 'City-Provinces'.

This law enabled the next step in the evolution of regional governance for metropolitan Amsterdam. The transformation of the ROA from a consultative group to a body that would make regional policy and have co-ordination powers was symbolised by its new name: Regionaal Orgaan Amsterdam. However, pending progress in the transition process prescribed by the Framework Law, the ROA remained a relatively weak organisation. Its offices in Amsterdam's City Hall are barely distinguishable from the adjoining suites housing Amsterdam's Economic Development Department. Its small (15-18 at its maximum) professional staff consists entirely of officials seconded from member municipalities, who continue to pay their salaries, and its only funding base is a ƒ 3,- per head levy, paid by its fifteen member municipalities. The disparity between any functional definition of the Amsterdam metropolitan region and ROA's boundaries is another problem (Van der Veer 1994), intensified by some communities' reluctance to participate, and others' early withdrawal.

But the emergence of metropolitan regional government in the form of City-Provinces, though endorsed by the political establishment and widely supported by the planning community, was not inevitable. Both its components-the formation of new metropolitan city-provinces, and the break-up of the dominant central cities-had opposition. In Amsterdam and Rotterdam opponents demanded referenda on the proposed changes, with results that stopped the trend towards formal metropolitan-regional government dead in its tracks.

This is the context for metropolitan-regional planning in the Amsterdam area today: a situation that is intrinsically problematic and unstable. But it is likely to last for several years, until the trend towards stronger metropolitan-regional governance reasserts itself. Meanwhile, it appears that intergovernmental co-ordination in the Amsterdam metropolitan region may retreat to its previous forms. The Province of North Holland seems poised to resume its traditional role in the scheme of intergovernmental relations, though implementation of its compact with Amsterdam may have introduced some irreversible changes. There are contradictory predictions about the future of the ROA: whether it will disappear entirely, or revert to its earlier consultative format as the IAO. Amsterdam's municipal administration will continue to enjoy its enhanced powers delegated by the Province, but the trend towards more independent sub-communities in the city has been halted, and may even be reversed.

Integrative planning and co-ordination

Regional-metropolitan planning in the Amsterdam area can only be appreciated in the context of the dynamics of emerging and retreating regional governance described above. To understand the regional planning process, it is useful to distinguish between five distinct levels at which integrative planning takes place. At the highest level is the regional and local interaction with national government; the next level involves the Province of North Holland and its relations with metropolitan and local agencies. The third level is the urban agglomeration or metropolitan region, for which the ROA has prepared a structure plan. Then there is integrative planning on an area-wide basis for special development areas such as Schiphol airport and its surroundings and the North Sea Canal corridor. Finally there is the significant element of regional-metropolitan planning in Amsterdam's own Structure Plan.

National-local interaction - The VINEX accords

The 1988 Fourth Nota reiterated previous national spatial policy: urban concentration in the Randstad and preservation of the Green Heart. Its stress on globalisation, economic competitiveness and regional differentiation had particular implications for the Amsterdam region. But subsequent discussion was overshadowed by a revolution in national housing policy, involving a significant reduction of the central government role. This included massive reductions in the budget for housing and more targeted social housing subsidies.

During debate on its adoption there was also the sense that changes were making the Fourth Nota increasingly obsolete. These included political changes, such as a new government, and environmental problems producing a resurgent environmental consciousness. This was the stimulus for the updating of national physical environmental policy presented in the Fourth Nota Extra (VINEX), which was adopted in 1992. Its focus was sustainable development, for which a new planning tool was the identification of ROM areas and key national strategic projects, including Schiphol national airport.

An important new implementation tool introduced under VINEX was the negotiated execution of national policy and resource allocations through intergovernmental covenants with local authorities. The VINEX covenants address the location of housing and the allocation of social housing subsidies to local governments, and the distribution of central government subsidies for major infrastructure development (Faludi and Van der Valk 1994: 217-23; Siraa, Van der Valk and Wissink 1995: 338-43). All these are issues of prime importance for regional land use and transportation planning.

The basic idea behind the VINEX accords is a 'stick and carrot' approach to implementing national policy. By agreement between the National Planning Service (which prepared the Nota), the Ministry of the Interior and the Housing Directorate of VROM, the Fourth Note Extra became the vehicle for linking housing subsidy with regionalisation. The 'carrot' for communities' agreement to regional government would be their participation in the allocation of national ministries' social housing, transportation and infrastructure subsidies.

The national Housing Directorate wanted to end local governments' dependence on housing subsidies, which traditionally covered the whole deficit of approved housing projects. The Fourth Note Extra proposed to substitute a global allocation for identified housing locations in each urban agglomeration: a 'lump sum' subsidy for the next ten years, which would cover about half the estimated cost shortfalls. This policy change implied cross-subsidies within municipalities (e.g. transferring some of Amsterdam's Land Utility's profits from business development to cover housing projects' deficits) and between communities in the region.

The case of IJburg, Amsterdam's latest expansion housing neighbourhood, illustrates the role of the VINEX accords in the area's metropolitan-regional planning. This potential location for 18,000 dwelling units was the topic of discussions with the national Housing Directorate back in 1978, when it was on Amsterdam's 'wish list'. In spite of its long-standing (and well founded) suspicion of Amsterdam's cost estimates, the Housing Directorate approved the IJburg project when it was advocated by Amsterdam and the ROA in the last round of VINEX negotiations (Korthals Altes 1994: 125-32). As a result, IJburg is an integral part of the housing proposals in the ROA and Amsterdam structural plans.

The Provincial role and its structure plan

The Dutch 'decentralised unitary state', based on the principle of co-government, gives the Provinces an important role in mediating central-local intergovernmental relations. One aspect of this role is the Province's preparation of a statutory land use and development structure plan, the streekplan, which can cover parts or the whole of a Province's territory (Bouwer 1994: 108; Wissink, Faludi and Lingbeek 1994: 178-80). North Holland, for the purposes of its structure plan, is divided into three parts, one of which covers much (but not all) of the area of the Amsterdam metropolitan region. This is the subject of the Amsterdam-North Sea Canal (ANZK) streekplan, which was completed in 1987.

The 1987 ANZK provincial structure plan is basically an integrated compilation dovetailing the implications of national policy for the region, as presented in the Fourth Nota, with local structure plans that were completed or (like Amsterdam's) in progress. Several developments have made this plan obsolete, but at the same time postponed its systematic revision. One is the changes in government policy, with the Fourth Nota Extra and negotiation of the VINEX accords. Another is the North Holland Province-Amsterdam compact, which substitutes Amsterdam's structure plan as the relevant streekplan for the city's territory, and the emergence of ROA as the designated metropolitan-regional planning agency.

Nevertheless, pending completion of the prescribed process leading to an independent city-province, the Province still had an important consultative role in approving the ROA structure plan. Its review, which at the time of writing was still in progress, was rather negative, and the Province withheld its approval. Among its objections were inconsistencies between the ROA structure plan and some of the VINEX accord provisions, and the apparent reluctance of ROA to set priorities in its plan for prospective residential development. The provincial planning unit also has the ongoing task of selectively amending and updating the 1987 streekplan, so that it could continue to perform its designated function. Such amendments include incorporating Amsterdam's approved housing expansion project of IJburg (Provincie Noord-Holland / Gemeente Amsterdam 1995).

Meanwhile, however, the Province's planning activities have again been overtaken by events: the referenda in Amsterdam and Rotterdam and their traumatic reverse of the expected trend towards regional-metropolitan government with virtual disappearance of the Province's role. As a result, an old idea is being revived: to replace the 1987 ANZK plan with a new provincial streekplan covering the whole southern half of North Holland, an area almost exactly matching the functional Amsterdam metropolitan region (with the exception of Almere). With the retreat of metropolitan regionalisation, the Province is likely to resume and even strengthen its traditional role of arbitrator of conflicts between local communities, and mediator between the national and local levels of government. In that case, preparation of the new provincial structure plan may become the main arena for anticipatory co-ordination of metropolitan regional development.

The metropolitan region - ROA and its structure plan

The ROA presented its preliminary structure plan (voorontwerp) in 1993. The final definitive plan (ROA 1995) included amendments in response to reactions to the preliminary plan from a wide spectrum of governments, public agencies, and organisations inside and beyond the ROA area. The plan tried to integrate other planning projects, such as the NZK Master Plan, the Schiphol ROM project, and the IJmeer housing development, as much it could. But because many of these projects were themselves in progress, complete integration was of course impossible.

ROA's regional structure plan was developed by its own small professional staff, aided (for particular elements) by consultants. However, in its planning the ROA drew extensively on the Province's and its member communities' resources: existing plans, ongoing planning efforts, (e.g. Amsterdam's structure plan, described below, and the Haarlemmermeer structure plan in development), and previous studies and analyses offering a wealth of data. The process of developing the ROA plan involved frequent and intensive interaction of ROA officials and planners with other members of the planning community, and a great deal of consultation, formal and informal, between ROA and other units and agencies of government.

Designed to be an integrative decision framework for land use and general spatial development, the Structure Plan offers relatively little detail regarding specific sectors. In terms of intersectoral co-ordination, it is intended to be complementary to more detailed sectoral plans: the regional Transportation and Traffic Plan and Housing Plan, a regional economic development strategy, and the regional Environmental Quality Plan.

The plan is based on analysis of regional spatial structure and trends, first addressing its urban and rural components separately and then integrating them into a comprehensive spatial framework. Its main objective, expressed as the plan's integrating evaluation criterion, is 'sustainable growth' (groei in duurzaamheid), where sustainability is defined as creating an environment that is liveable in the long-term. The plan identifies its basic constraints: development limits related to Schiphol airport's noise pollution, existing and prospective infrastructure corridors, and conservation of critical areas. These are land with serious soil pollution problems, or areas valued as agricultural, cultural or natural resources. Sectoral trends are analysed and options are identified in housing, traffic and transportation, services, environmental quality and rural areas.

A major policy element of the ROA plan is the location of potential housing sites in the region. These were also the subject of negotiations between ROA and the national Housing Directorate leading up to the VINEX accords, which paralleled development of the ROA plan. But, while the national ministry initially wanted to bargain these accords exclusively with ROA, it soon found ROA too weak, as a governmental unit that was still more emergent than real, to make the necessary local commitments, and instead the region's major housing expansion communities (bouwgemeenten) became the government's negotiating counterparts. As a result, Amsterdam played a major role in this process (Cornelissen 1994: 145-8), as we will see below.

The plan's housing element addresses the question of how and where to accommodate the 100,000 dwelling unit (DU) demand estimated in VINEX-related projections for 1995 to 2005. The basic policy options were to emphasise densification and infill housing within Amsterdam, or to disperse housing construction throughout the region. The ROA's conclusion was to do both.

The region's potential housing capacity is identified at 100-120,000 DUs distributed over 31 sites. 21-28,000 DUs are on about 31 ha. of densification locations, with an average proposed density of 80DUs/ha, within existing built up areas. The remainder, amounting to at least 80,000 DUs, are proposed for 'greenfield' housing expansion locations. At an average density of 33DUs/ha, these sites will consume about 2400 ha. of currently open land and water. The plan tries to limit land consumption by recommending the highest possible densities, but this advice can be problematic.

One such case is the IJburg housing expansion area. No one considered this an ideal location: its high development costs would need to be offset by building to higher densities, which would make IJburg less attractive to exactly the higher-income and status population Amsterdam wants to attract. Its offshore location also dooms IJburg to an almost inevitable transportation bottleneck in its link back to the mainland.

An alternative location which has long been in competition with IJburg for Amsterdam's overflow housing is the Bovenkerkerpolder, on the edge of the southern 'lobe' of the Amsterdam conurbation and north of the village of Uithoorn, which is already within the 'Green Heart'. This site suffers from neither of IJburg's above disadvantages, but was rejected for other reasons. Urbanising this area would penetrate the 'Green Heart', which national policy forbids, reducing this site's feasibility. But it is also a much less desirable location in Amsterdam's view, because it is outside the city boundaries. Ultimately, in the ROA plan as in the VINEX compact, the Bovenkerkerpolder is retained as a possible reserve site for post-2005 housing.

The ROA structure plan distributes future housing production according to the VINEX accords, and envisages full development of all the identified expansion housing locations. It also proposes intensive development of all prospective employment opportunities: these include the Schiphol international airport area, the North Sea canal and port, and a number of identified regional industrial parks. The plan also identifies missing and deficient elements of the regional transportation network, including various modes of public transportation, highway links, and bicycle routes, which are included in the plan's a summary identification of regional spatial projects.

In its implementation strategy the plan presents itself as an advisory frame of reference (like the streekplan) for the region's municipal statutory land use plans (bestemmingsplannen), leaving local governments due leeway for interpretation. Among its designated implementation tools are the VINEX accords. These will relate future housing development to prescribed VINEX locations and their associated housing subsidies, dividing ƒ 4.2 milliard over the next decade between Amsterdam, Haarlemmermeer, Purmerend and Zaanstad. Implementation of the planned transportation and infrastructure projects is also handled through the relevant VINEX accords. In its other sectors (environment, services and rural areas) the plan leaves implementation of its provisions (through mandated planning and appropriate budget allocations) to be detailed in the relevant sectoral plans (Schuiling 1994; ROA 1995).

Integrated Area Planning

Integrated planning in the Amsterdam metropolitan region is not confined to regional plans. Several areas have been identified as subjects for integrated area planning. Planning these projects is parallel to the other regional planning efforts described here, and there is considerable interaction between them. This takes a variety of forms: issue-related or functional special task forces linking the various planning teams, individuals filling multiple formal roles involving them in several levels of planning, and informal consultation and information exchange. Accordingly, the area-wide level of planning is an integral part of this complex system of metropolitan-regional planning.

The integrated planning of the area around Schiphol international airport is an example. This area was designated as a ROM area in view of its unique characteristics. These included problems related to air traffic: noise and air pollution, soil contamination, and conflicting demands on limited space by transportation requirements, airport related commercial and industrial development, housing supply and open space. The area's opportunities include its employment potential and its role in the national economy. The airport's prospective contribution in 2015 is estimated at 3.2 per cent of the Netherlands' GNP, and its performance is critical for the country's role as a major European distribution centre. To maintain Schiphol's competitive position its expansion is essential, but the ensuing safety problems were highlighted by the 1992 crash of a cargo plane into a densely populated Amsterdam suburb.

In a compact they signed on 11 September 1989 three national ministries, the Province of North Holland, the communities of Amsterdam and Haarlemmermeer, the Airport Authority and KLM became the participants in a special organisation set up for the purpose of integrated ROM planning: Project Mainport Milieu Schiphol (PMMS). The PMMS was run by a steering group of politicians and officials representing the participating organisations, which met several times a year. A full-time Project Director, seconded from the Ministry of Transportation and Public Works (TPW), headed a management group consisting of senior professionals from the respective agencies who were involved in execution of the project, which met more often.

PMMS was staffed by a small temporary multidisciplinary planning team backed up by the appropriate technical units of the participating agencies, and supported where needed by outside consultants. Subteams were responsible for sectoral planning in areas such as transportation and environment, and study groups were formed for special analyses such as economic and environmental impacts. The project also commissioned fifteen research studies from consultants.

An important contribution was the recently published Schiphol Airport Master Plan, which proposed major expansion of the airport to accommodate prospective traffic, including addition of a new fifth runway. One of the main questions the PMMS plan had to address was the level of proposed capacity increase of the airport, and whether the additional runway was necessary. Other issues included the airport's links with the High Speed Rail network, and its future operating characteristics which would determine the location of flight-related noise contours.

During the plan's preparation there was intensive lobbying. The small town of Aalsmeer which was most affected by noise and air pollution wanted to be represented on the Steering Group, but it was not given official status, and had to lobby just like other interest groups. Residents' groups and environmental organisations formed a special interest organisation, the Forum for the Living Environment in the Schiphol Airport Region, which opposed airport expansion. It proposed limiting future capacity and rotating the use of the existing runways rather than adding another one, and demanded restrictions on night flights.

The Draft Action Plan issued in December 1990 advocated controlled growth of the airport. It proposed expansion of Schiphol's capacity to 40m. passengers and 3m. tons of freight by 2015, by constructing the fifth runway, allowing future residential construction only in the area outside the future maximum noise contours, and installing acoustic insulation in existing dwellings. To reduce aviation related impacts, the plan proposed major expansions of rail service, diverting as much short-haul air traffic as possible to the High Speed Rail line, and modifying operating patterns to minimise use of the most noise-sensitive runways and limit night flights.

Unresolved issues included the regulation of night flights in the absence of national noise standards. Public debate and bilateral negotiations between the Ministry of VROM and affected parties continued until signing of a policy covenant in April 1991 to implement the plan, including its development into a national 'Key Planning Decision' (PKB). An extended consultation procedure was undertaken with two groups. The first was governmental bodies: the ROA, adjacent municipalities, special districts, etc.; the second comprised non-governmental interests: environmental and residents' organisations, NGOs, and associations and firms from industry and commerce.

In December 1993 the government published the first part of its PKB, which included its decision to construct the fifth runway. In what was again an exhaustive public participation process (in the course of which 12,000 objections were received and responded to) this was vehemently opposed by residents' groups and environmental organisations, which, anticipating this outcome, had earlier already withdrawn from the formal consultation procedure. But the final PKB (issued in February 1995) essentially confirmed the PMMS Action Plan.

Staged implementation of the Action Plan included the Province's amending its statutory ANZK streekplan to incorporate relevant land use and development provisions (Provincie-Noord-Holland 1995), and amendment of the Civil Aviation Act to formalise new permitted noise zones. It also includes development of airport related economic functions, such as creating 75 ha. of industrial parks for transportation and distribution firms. For this purpose a public-private partnership was formed, the Schiphol Area Development Co., in which the Province, the municipality of Haarlemmermeer, the Schiphol Airport Authority and the National Investment Bank (which is providing the financing) are partners (Ministry of Housing, Spatial Planning and the Environment 1995: 50-9; Van der Berg and Schaafsma 1993, 1996; Wissink n.d.).

The Amsterdam region and the Amsterdam Structure Plan

Amsterdam's own planning is an important element of the metropolitan-regional planning system, for two reasons. One is that Amsterdam is a dominant part of the area; this alone makes Amsterdam's planning significant for its metropolitan region. But the other reason is that Amsterdam has never constrained its planning to its municipal boundaries. Amsterdam's politicians and planners always regarded its region as its natural planning area, both for the purposes of learning and analysis, and as the subject territory for planned public intervention. Consequently, the story of Amsterdam's planning is not only local: it is also part of the story of metropolitan-regional planning and development.

Structure planning in Amsterdam is an ongoing process, with plans produced roughly at five year intervals. The 1974 Structure Plan was no exception to its predecessors in its regional orientation. Its focus on central city revitalisation and renewal was already in reaction to the perceived effects on Amsterdam of then prevailing regional development policies which located most of the city's overspill housing in regional 'growth centers' (Wallagh 1994: 147-57, 181-2, 191-3; Faludi and Van der Valk 1994: 163-70).

'Compact city' was the core concept of the 1985 Structure Plan, entitled 'Focus on the City' (De Stad Centraal). What was foreshadowed in the 1974 Structure Plan became a veritable revolution in the 1980s. Amsterdam's new planning doctrine was not pioneered by its planning establishment, which still adhered to the national and regional policy of 'concentrated deconcentration', and wanted to assure continuity of some of the major strategic infrastructure projects which others saw as depopulating the inner city. Rather, the planning revolution was stimulated by a political turnover in city government, which was taken over by a faction of the Labour Party (PvdA) led by local politicians who began as community activists (Faludi and Van der Valk 1994: 186-7; Wallagh 1994: 214-31).

Amsterdam's 1985 Structure Plan is significant because it set the parameters of the 'compact city' planning doctrine and its accompanying spatial organisation principle, which frame its planning and regional development policies to this day. These include focusing development within the core city, with priority for densification and infill housing before considering expansion sites to fill unmet housing demand. Proximity to employment opportunities and access by public transport became key criteria for evaluating areas' development potential.

Renewed attention to urban revitalisation and economic development took several forms. The plan included proposals for more market housing, and several major infrastructure projects, such as the completion of the Ring Road. A proposed tunnel under the North Sea Canal to strengthen the link with suburban communities to the North (which, however, was not carried out) is also evidence of the plan's regional orientation. Several major development projects, to be implemented by public-private partnerships, included mixed use construction and adaptive rehabilitation of the harbourfront on the 'North Axis' along the IJ waterway, and 'edge city' type subcentres (including the Sloterdijk teleport) on the 'South Axis'. The plan's spatial organisation principle was still faithful to Amsterdam's tradition: urbanised densely built-up 'lobes' separated by wedges of green space that were linked to the open hinterland (Faludi and Van der Valk 1994: 186-92; Wallagh 1994: 238-42, 250-7).

This plan was drawn up in Amsterdam municipal administration's Planning Department (DRO) by a special task force (Werkgroep Struktuurplan). The plan was developed using the administration's rich ongoing data base, and was preceded by a considerable number of interim products: analyses, reports, and draft policy proposals (Wallagh 1994: 242-8). This would be the process too for the next (1991, 1996) Structure Plans.

The 1991 Structure Plan essentially follows the same planning doctrine, with some changes in format (Wallagh 1994: 311-2). One change is increased emphasis on environmental policy and impacts (DRO 1994a: iii). Subsequent work toward the next (1996) Structure Plan involved serious debate on whether to retain the 'compact city' concept as the plan's framing principle, or whether this doctrine had run its course (primarily due to exhaustion of the city's potential housing capacity). Finally, in the absence of any consensus on a replacement, the 'compact city' was reiterated as the framing principle for Amsterdam's spatial development policy in the 1994 Draft Structure Plan (DRO 1994b: 9).

Like its predecessors, this plan was prepared by the Structure Planning unit in the DRO, in a highly interactive process which included publication and debate on several preceding reports. One was the 1993 Progress Report on implementation of the 1985 Structure Plan, which also discussed problems addressed in the 1991 Structure Plan: housing, employment opportunities, mobility and environment. This Report included the implications of the (1992) Risk Analysis of Projects for Spatial Renewal in the course of formulating the principles of a new physical development policy (Wallagh 1994: 315-7).

Another lead up to the 1994 Draft Structure Plan was a statement of development policy presented in the (1993) Nota: Amsterdam Toward 2005. This Nota illustrates the interaction between Amsterdam's structure planning and the metropolitan-region's planning. It is Amsterdam's input into the VINEX negotiations that were taking place with ROA. It also shows the shift from basic planning and spatial development policy, which was by now quite well articulated and generally accepted, to implementation strategy, in which the VINEX accords and their related subsidies played an important part.

The Nota's development priorities also show its metropolitan orientation. Besides Amsterdam's own infill housing and its (partly extraterritorial) expansion area of IJburg, the Nota identifies several critical regional infrastructure projects, including construction of the first phase of the High Speed Rail connection between Schiphol and Amsterdam, and of the Westrandweg highway connecting Schiphol and the western harbour (Wallagh 1994: 326, 332-3).

'Amsterdam Open City' is the title of the 1994 Draft Structure Plan, which essentially details and formalises the 1993 Nota's urban development policies. Though it did not deploy any very sophisticated models or methodologies, the DRO's Structure Planning unit produced a highly competent and very readable plan. The plan opens with a summary of the city's spatial development policy, which again demonstrates its regional orientation.

The expressed priorities of Amsterdam's 'compact city' policy are core city revitalisation, accommodating regional growth and enhancing regional linkages, including development of the IJ waterfront and IJburg. Another priority is strengthening the Randstad's international position through co-operation in regional development projects: the North-South rail line, Schiphol development, High Speed Rail links, and strengthening of the West harbour.

A regional orientation shows in the plan's analysis, which reviews the conurbation's international competitive position and the implied spatial policy choices leading to the 'compact city' policy. Among the plan's sectoral elements with regional implications are spatial structure, with policies of densification and diversity of landuses, green areas, prescribing protection of the open 'wedges', and population and housing policy which identifies regional residents among its target groups. Housing implementation measures include the VINEX housing program, and infrastructure and transportation identify several projects that extend into the region well beyond the city boundaries (DRO 1994b).

Sectoral planning and implementation

Housing

Though it is a distinct sector, the fact that housing is one of the principal concerns of regional integrative planning is obvious from the description above. The location, mix, and financing of housing are major elements of Amsterdam metropolitan-regional planning. Regional housing policy has developed incrementally, based on prior demand estimates and previously proposed projects to supply projected housing needs. An example is the long running debate on IJburg.

The role of housing as the leading sector in Amsterdam's metropolitan-regional planning was institutionalised in the VINEX negotiations and accords. These, in fact, determined the main elements of the spatial form that development of the metropolitan region would take, which was only formalised in the ROA's and Amsterdam's Structure Plans. At the same time, prior regional plans themselves (the 1989 Amsterdam Structure Plan, preliminary ROA reports, and the 1989 Provincial ANZK streekplan) set the framework for these negotiations.

The VINEX negotiations opened in January 1991 with the national government's 'offer' based on the 1990 VINEX national 'key planning decision' (PKB). Minister Albers of VROM opened discussions with Amsterdam ex-mayor Van Thijn, the ROA chairman, but negotiations soon bogged down on inherent uncertainties in the government's proposal. These included estimates of land acquisition and development costs, prospective market demand, and potential private participation, leading ROA to project the risk of a possible ƒ 540m. deficit. ROA's counterproposal to guarantee half of any eventual shortfall if the government agreed to cover the other half was unacceptable to the Ministry, which wanted a firm implementation commitment from its local government counterparts.

Minister Albers dissolved the deadlock by incorporating these risk contingencies in the government's offer. If ensuing developments were positive, the ROA guarantee would be sufficient. If they were negative, raising expectations of larger deficits, the accords could be renegotiated in 1999, including reconsideration of alternative, cheaper, housing locations. This would mean, for example, substituting the Bovenkerkerpolder as Amsterdam's major expansion site for phase II of the much more costly IJburg project. These negotiations concluded with a cabinet resolution (PKB Spatial Policy 1991) and a formal agreement between the parties.

But implementation soon ran into problems. Reassessment of the risk factors included rising land development costs and growing demand. By 1994 the accords' need estimates were overtaken by actual demand and steeper trend projections, raising the spectre of a major housing shortage. This led to a reassignment of housing allocations: Amsterdam's was raised from 87,000 to a minimum of 100,000 DUs. It also sparked reconsideration of some expansion housing locations and additional infill housing in Almere and Haarlemmermeer (Wallagh 1994: 330-2). These were included in the 1994 Amsterdam Structure Plan and the 1995 ROA Structure Plan.

The IJburg project illustrates the development of a major housing area in the context of the region's metropolitan planning. The first proposal was by an interorganisational group (STAW) in the Amsterdam municipal administration as a development for 18,000 DUs on reclaimed land. The adjacent community of North Diemen objected, seeing the project as conflicting with its own development plans and expecting negative impacts from its proposed highway connection. But IJburg became an integral part of Amsterdam's 'compact city' policy in its 1985 Structure Plan-only one instance of this plan's extraterritorial reach. Subsequent negative reactions from several neighbouring communities suggested referring the proposal to the (in fact defunct) Stadsrandcommissie.

Now the Province also got involved: though its plans had indicated IJmeer as a possible housing location, the Provincial Planning Commission saw problems of negative environmental impacts. In response, Amsterdam began intensive lobbying to obtain the provincial and national governments' consent, which was essential both for legal conformity and to ensure needed funding. 1987 saw Amsterdam's first success: the Province included IJburg in its amended streekplan.

But the Ministry of VROM was still considering suitable housing sites in the region. To facilitate its choice of optimal housing locations, it initiated a 'Tripartite Consultation' process, creating the Tripartite Working Group Urbanisation Amsterdam in 1987. This interorganisational group, representing the national government, the Province of North Holland, and the municipality of Amsterdam, identified priority regional housing locations, including the IJburg site. Partly, this was the result of a coincidence of interests between Amsterdam and the government: both supported large-scale housing production.

In 1991 IJburg's future seemed assured when it became part of the VINEX accords. Meanwhile, Amsterdam proceeded with more detailed planning, including revision of its statutory bestemmingsplan. Statutory revisions also required shifting the municipal boundary to include phase II of the IJburg development. This would annex part of the neighbouring community of Muiden, but it was approved by the Province in spite of Muiden's objections.

Problems on two fronts with IJburg's planning and development have delayed implementation. On the environmental front there has been opposition by 'green' interests, who projected severe negative impacts on the IJmeer's fragile ecology. Environmental Impact Statements (MERs) were required, demanding time and resources. The MER for the project's second phase included an evaluation of alternative configurations of the reclaimed land area to minimise environmental impacts (DRO 1995), but 'green' opponents may still hold up IJburg's development by demanding agreement by referendum.

On the economic front all is also not clear sailing. The IJburg development will be very expensive because of its land reclamation and infrastructure costs, and its concentration on market housing limits its feasible densities. Nevertheless, the project seems to have economic potentials that look tempting for possible public-private partnerships. One development syndicate (Consortium Nieuw-Oost) presented an unsolicited proposal in 1990, and in 1993 the municipality was negotiating with another, the Waterstad Group, to form a partnership to implement the project's first phase (Cornelissen 1994; Korthals Altes 1994; Schuiling 1994; De Vries 1994).

Transportation and infrastructure

Regional integrative planning, as described above, addresses these sectors too. For example, the VINEX accords included a significant transportation and infrastructure component, which was negotiated on the government side by the Ministry of TPW. In a recent policy change, the Ministry had decided to emphasise public transit and retrench its highway program; this was also consistent with the National Planning Department's (RPD) policy.

For the Ministry of TPW, the VINEX procedure and its subsidy 'offer' were a major break with its previous practice of funding individual projects following applications from local governments. After some persuasion by the RPD, the Ministry of TPW also participated in the VINEX subsidy offers to the four major conurbations, which included Amsterdam. But (in contrast with the housing sector) the transportation and infrastructure offer only gave an indication of the dimensions of possible subsidies; the actual funding commitment would still be on the basis of specific project review.

There were two reasons for this procedure. One was that it was prescribed by law. The other was that, compared to the housing sector, where the market still provided a significant part of the supply, transportation and infrastructure involve much higher levels of government funding. This, coupled with the complexity and risk involved in major strategic projects, made funding commitment by project imperative (Korthals Altes 1994: 132-4). Nevertheless, the VINEX accords identified the major regional transportation and infrastructure elements which were approved in principle, and these were included in the ROA and Amsterdam Structure Plans.

But a separate plan for transportation is also required. This is a reflection of the fragmented institutionalisation of these sectors. Transportation is the domain of a special purpose 'district', but during the period covered here the ROA also took over this district's planning functions. Consequently, though transportation and traffic planning has been integrated into the regional Structure Plan, the ROA is preparing a more detailed Traffic and Transportation Plan to comply with statutory and funding requirements.

There are also various agencies for other parts of the regional infrastructure. The responsibility for sewerage and water is divided. Municipalities are charged with local sewage collection and transfer, but water management is a separate domain. The Amsterdam metropolitan region is divided into three Water Districts (Waterschappen), which are responsible for the quantity of water and its movement through the area's waterways, and a separate agency handling wastewater and sewage treatment. In fact, water management is a separate and highly institutionalised sector in the Netherlands. Co-ordinating water management with spatial-physical planning in an integrated system of metropolitan-regional planning is a problem that is not unique to the Amsterdam area, and will be discussed below.

Environment

Environmental policy is an increasingly important aspect of regional-metropolitan planning in the Netherlands. The ROA Structure Plan identifies environmental quality as one of its main goals, and in both the ROA and the Amsterdam Structure Plans environmental constraints are one of the major determinants of the proposed regional spatial structure.

Integrating physical planning with environmental planning and policy has been a major recent concern. This is not limited to cross-sectoral co-ordination, but also addresses intersectoral integration in the fragmented domain of environmental policy itself, where separate agencies are responsible for managing and monitoring land, air, and water. One way in which integrated planning has been attempted invokes a common territorial base, and sets up special integrated projects for identified ROM areas. The integrated Action Plan developed for the Schiphol area, described above, is an example of this approach.

The other attempt at integration has taken the form of requiring a distinct and intrasectorally co-ordinated 'stream' of environmental planning and policy at each level of government. At the national level this is the National Environmental Policy and Conservation Plan, which is also developed in interaction with and conformity to national spatial planning and development policy as expressed in the Notas on spatial order. This plan, for example, identifies nature conservation areas of national importance, and (together with the Fourth Nota) designated the ROM areas for integrated physical-environmental planning.

The next level of intra- and intersectoral integration is the Province, where the Provincial environmental policy plan must be co-ordinated with the relevant streekplan or plans (Brouwer 1994: 108-10). In the Amsterdam metropolitan region, the regional environmental policy is still in preparation as a supplementary element to the ROA Structure Plan. Here, as elsewhere, the problem arises of co-ordinating between statutory planning processes involving environmental policies and plans and spatial-physical development plans, often prepared by different agencies and always completed at different times. A legal-regulatory device that has been developed to provide a stimulus to enhance such co-ordination is the 'cross-over principle', which imposes the burden of ensuring consistency on the 'lagging' sector (Wissink, Faludi and Lingbeek 1994: 185-6).

Discussion and Conclusions

Amsterdam's regional planning, it turns out, is noteworthy for its complexity. The Dutch establishment's long-standing concern with the structural organisation of metropolitan-regional governance is evidence of a widespread perception that this is a problem. Nevertheless, at least to outside observers, the Netherlands present an image of a society that is quite systematically planned and well co-ordinated, and the reality of its developed and built environment attests to some successes in Dutch planning practice.

That is one reason for this study: the sense that the Amsterdam case might be an exemplar for metropolitan-regional planning and intergovernmental co-ordination, in the absence of institutionalised metropolitan-regional governance. What this exploration of the actual guidance of the development of Amsterdam's metropolitan region reveals, however, is that whatever successes are attributable to its planning cannot be explained by the formal structure of the planning system.

The quasi-hierarchical interaction between the three levels of government in the statutory system of spatial-physical planning does provide a co-ordinating framework. But the statutory spatial planning system is geared primarily to vertical co-ordination and formal public consultation. It is recognised as inadequate to address issues involving interdependencies between metropolitan-regional communities, which demand some arena of metropolitan-regional planning and horizontal intergovernmental co-ordination. A sense that the traditional statutory spatial planning system is also poor at intersectoral co-ordination is revealed in recent efforts at reform: new forms of integrated area-wide planning, attempts at forming metropolitan-regional arenas for intersectoral interaction, and rules for intersectoral planning co-ordination at the national and provincial levels (Wissink, Faludi and Lingbeek 1994; Brouwer 1994).

Formal and institutionalised metropolitan intergovernmental co-ordination, in the form of ROA as a unit of emergent regional governance, and ROA's initiatives in metropolitan-regional planning, also cannot be significant factors in evaluating Amsterdam's regional planning performance. This might have been different if we were looking at this region in (say) the year 2005, and if the apparent evolution of metropolitan-regional governance into city-provinces had run its course. But it was 1996, and ROA had only taken the first steps of its transformation from its previous identity as a more formalised consultative arena. There is ample evidence of ROA's weakness and its relatively small contribution to metropolitan planning and co-ordination of the Amsterdam region: its low organisational autonomy, its tentative Structure Plan, and its role in the VINEX negotiations. The conclusion is compelling that metropolitan Amsterdam was still a long way from regional governance.

Just as metropolitan governance can be confidently excluded as a factor in accounting for the performance of Amsterdam metropolitan-regional planning, so can another possible explanation: planning doctrine. The existence of a planning doctrine-a conceptual 'frame' for an area's spatial development that has been adopted by the relevant policy and planning communities-can enhance the quality of planning and positively affect planning performance (Alexander and Faludi 1996). Dutch national spatial-physical planning is the classic case (Faludi and Van der Valk 1994). But here there is no planning doctrine; indeed, the preconditions for a planning doctrine for the Amsterdam metropolitan region do not even exist.

There is another explanation which could account for the successes (such as they are) of planning in the Amsterdam metropolitan region. This factor is not limited to the area of our discussion, but applies to Dutch planning in general. I am referring to the public role in development, which is unique to the Netherlands (Alexander 1988). Perhaps the fact that government is not only a planner-regulator, but also the initial land developer, makes the built environment conform to plans more than it would otherwise, in spite of the fragmentation of metropolitan-regional planning that we have seen.

There are also two ways in which metropolitan-regional planning in the Netherlands has addressed and, to some degree surmounted, its complexity. This may be another reason for the positive results shown in the planned environment, in the face of apparent confusion in the planning system. One is how the planning process is run: the extent of formal and informal consultation. The other is how the planning system is institutionalised (including its consultative component): the widespread and deliberate creation and use of IOC systems and structures.

Consultation is an integral part of the Dutch planning process. This is not only the formal consultation (in the form of hearings, receiving objections, etc.) required in most countries, which may be little more than token participation. Formal intergovernmental consultation and interaction in the course of developing and finalising plans is also a legal obligation in the Netherlands' various 'vertical' systems of planning and sectoral co-ordination: spatial-physical planning, environmental planning and policy, and transportation planning. This type of interaction is much more prominent in Dutch political culture than elsewhere: the National physical-environmental Notas, for example, are actively debated in parliament and formally adopted by the government, as are provincial plans as well, while local planning issues are also the subject of lively political discourse. Public participation requirements are also strictly observed, and actively utilised (as for example in the Schiphol case described above), though opinions vary about their impact on actual decisions.

But besides these forms of consultation, there is active interagency and interorganisational consultation in all phases of the planning process. This includes formal and informal interactions. Formal arenas for consultation and information exchange are ubiquitous: ad-hoc task forces, standing committees, working groups, steering committees, even consultative organisations (such as the IAO). As a form of IOC structure, they will be discussed below. These are the locus of intersectoral and 'horizontal' co-ordination, complementing the prescribed 'vertical' co-ordination built into the statutory planning systems. Supplemented by informal interpersonal interaction among members of quite a well-knit planning-policy community, consultation is an important explanation for how Dutch planning surmounts fragmentation and complexity to arrive at decisions by consensus.

The structural aspect of consultation is the wealth of IOC structures in which interorganisational interactions are institutionalised. These range from interorganisational groups (probably the most widespread) to co-ordinating units (Alexander 1995: 47-75), and include ad-hoc issue-based or project-related and standing or permanent IOC structures.

Amsterdam's municipal administration (involved in metropolitan-regional planning in many ways including its Structure Plan) offers many examples of interorganisational groups. These include interdepartmental working groups routinely set up for projects involving several sectors, and the interagency steering committees (stuurgroepen) covering intersectoral 'facets', which meet regularly for information exchange, consultation, and administrative coordination.

Formal interorganisational groups are also routinely set up as ad-hoc consultative arenas to resolve planning and policy issues affecting diverse interests-essentially planning as anticipative co-ordination. The Tripartite Working Group Urbanisation Amsterdam, described above, is an example More rare, but also instituted where necessary, is the standing interorganisational group institutionalising consultation and some co-ordination between a set of interdependent organisations. The Stadsrandcommissie of the late 50s and early 60s, and the more recent IAO which evolved into the ROA, were such groups.

Finally, organisations are created as co-ordinating units. These may be for a special purpose, such as the PMMS set up to integrate planning for the Schiphol area. Or they may be charged with ongoing co-ordination of their interorganisational network's relevant decisions and activities. The ROA is an example of the latter, where the participating members are the metropolitan region's communities. It remains to be seen whether this degree of regional integration is acceptable. As of this writing, it appears that less institutionalised and more consultative forms of co-ordination are perceived as more compatible with traditional Dutch local autonomy.

Consultation and its institutionalisation in IOC structures demand time, effort, and resources. In the Netherlands governments and planners are willing to invest these. They apparently have the justified expectation that they will be rewarded with more effective planning that overcomes the obstacles of fragmented institutionalisation and complexity to reach consensus on better decisions. Perhaps this is why the Amsterdam metropolis is indeed an exemplar of successful metropolitan-regional planning.

References

Alexander, E.R. (1998) 'Doing the impossible: Notes for a general theory of planning', Environment and Planning B: Planning & Design (forthcoming).

- (1995) How Organisations Act Together: Interorganisational co-ordination in theory and practice. Newark NJ/Montreaux: Gordon & Breach.

- (1988) 'The Netherlands' unique planning system: A visitor's comparisons', Rooilijn 21: 145-50.

- and A. Faludi (1996) 'Planning doctrine: Its uses and implications', Planning Theory 16: 11-62.

Barlow, I.M. (1991) Metropolitan Government. London: Routledge.

Berg, M. van der, and M. Schaafsma (1996) Planning Schiphol Airport: Planning between governments, entrepreneurs and citizens. 14/6/96.

- (1993) Schiphol: Development of the airport and the roles of various parties. 14/4/93.

Binnenlandse Bestuur (1995) articles 13/1/95, 16/6/95, 12/1/96.

Bos, C.A., J. Buit, K. Dobben and J. Walen (1988) 'Successen en mislukkingen in de Nederlandse ruimtelijke ordening', Planologische Studies 7. Amsterdam: PDI - University of Amsterdam.

Bouwer, K. (1994) 'The integration of regional environmental planning and physical planning in the Netherlands', Journal of Environmental Planning and Management 37, 1: 107-16.

Cornelissen, J. (1994) 'Nieuw Oost en nieuw bestuur'. pp. 139-50 in K. Bovenberg, W. Korthals Altes and A. van der Valk (eds.) Ruimte voor Regionaal Bestuur? Planologiese opstellingen over bestuurlijke samenwerking in de regio Amsterdam. Amsterdam: AME - University of Amsterdam.

DRO Dienst Ruimtelijke Ordening (1996) 'Amsterdam Gedeeld Geheel: naar een dynamischer struktuurplanning', Working paper: Structure Plan Unit. Amsterdam: August 1996.

DRO Dienst Ruimtelijke Ordening (1995) Milieu Effectrapport IJburg - Tweede Fase: Samenvatting.

DRO: Dienst Ruimtelijke Ordening (1994a) A City in Progress: Physical planning in Amsterdam. Amsterdam: Stadsdrukkerij.

DRO Dienst Ruimtelijke Ordening (1994b) Amsterdam Open Stad Ontwerp Struktuurplan 1994. Amsterdam: Stadsdrukkerij.

Evans, R. (1994) 'Birmingham'. pp. 68-86 in A. Harding, J. Dawson, R. Evans and M. Parkinson (eds.) European Cities Towards 2000: Profiles, policies and prospects. Manchester/New York: Manchester University Press.

Faludi, A., and A. van der Valk (1994) Rule and Order: Dutch Planning Doctrine in the Twentieth Century. Dordrecht: Kluwer.

- (1991) 'Half a million witnesses: The success (and failure) of Dutch urbanisation strategy' in A. Faludi (ed.) Fifty years of Dutch national physical planning (special issue) Built Environment 17: 43-52.

Frisken, F. (1993) 'Planning and servicing the Greater Toronto area: The interplay of Provincial and municipal interests.' pp. 153-204 in D. N. Rothblatt and A. Sancton (eds.) American/Canadian Intergovernmental Perspectives. Berkeley: University of California, Berkeley - Institute of Governmental Studies Press.

Gelfand, M.J. (1993) 'Development policy in metropolitan Boston'. pp. 53-110 in Rothblatt and Sancton (eds.) op.cit.

Hajer, M.A. (1997) 'The Netherlands: Strategic planning in crisis.' Planning Theory (forthcoming).

Harding, A. (1994) 'Amsterdam and Rotterdam'. pp. 18-44 in A. Harding et al. (eds) op.cit.

Het Parool (1995) articles 20/4/95, 13/5/95, 18/5/95, 20/5/95.

Hemmens, G.C., and J. McBride (1993) 'Planning and development decision making in the Greater Chicago region'. pp. 111-52 in Rothblatt and Sancton (eds.) op.cit.

Korthals Altes, W. K. (1994) 'De convenanten: Het Rijk stimuleert regiovorming'. pp. 124-38 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

Kunzmann, K.R., and M. Lang (1994) 'Frankfurt'. pp. 88-105 in A. Harding et al. (eds.) op.cit.

Levine, M.A. and J. van Weesep (1988) 'The changing nature of Dutch urban planning'. Journal of the American Planning Association 54, 4: 315-23.

McDowell, B.D. (1988) 'Regional Planning Today'. pp. 133-65 in F.S. So, I. Hand and B.D. McDowell (eds.) The Practice of State and Regional Planning. Chicago: APA.

Ministry of Housing, Spatial Planning and the Environment (1995) The Gordian Knot Untied: Integrated regional planning in the Netherlands. The Hague: Ministry of VROM.

Needham, D.B., B. Kruijt and P. Koenders (1993) Urban Land and Property Markets in the Netherlands. London: UCL Press.

Provincie Noord-Holland (1995) Partiële herziening Streekplan ANZKG: Haarlemmermeer/Schiphol. Haarlem.

Provincie Noord-Holland/Gemeente Amsterdam (1995) Streekplan Noord-Holland Zuid: Herziening grondgebied Amsterdam - 2e wegontsluiting IJburg. Haarlem.

ROA Regionaal Orgaan Amsterdam (1995) Regionaal Structuurplan 1995-2005. Amsterdam: ROA.

Rooij, E-J. de (1994) 'Beleidsvrijheid ruimtelijke ordening Amsterdam: Eeen overgangsregeling naar een 'Provincie nieuwe stijl'?'. pp. 90-102 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

Schön, D.A. (1997) 'Planning - Doctrine or Dialectic'. Planning Theory (forthcoming)

Schuiling, D. (1994) 'Zoeken naar de ruimte voor woningbouw binnen ROA: Van compacte stad naar compacte regio'. pp. 75-89 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

Shetter, W.Z. (1986) The Netherlands in Perspective: The Organisation of Society and Environment. The Hague: Martinus Nijhoff.

Siraa, H.T., A.J. van der Valk and W.L. Wissink (1995) Met het Oog op de Omgeving: Een geschiedenis van de zorg voor de kwaliteit van de leefomgeving - Het ministerie van Volkshuisvesting, Ruimtelijke Orgening en Milieubeheer (1965-1995). The Hague: SDU Publishers.

Thomas, T.E. (1993) 'Edmonton: planning in the metropolitan region'. pp. 245-82 in in Rothblatt and Sancton (eds.) op.cit.

Valk, A. van der (1994) 'Tussen servet en tafellaken: Het Regionaal orgaan Amsterdam op weg naar de status van provinsie nieuwe stijl.' pp. 28-44 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

- and G. Wallagh (1994) 'Amsterdam en haar buuren: de geschiedenis van het bustuurlijk overleg vanuit een planologisch gezichtspunt'. pp. 11-27 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

Veer, van der, J (1996) 'Metropolitan government in Amsterdam and Eindhoven: A tale of two cities'. Paper presented at the International Geographic Congress, The Hague, August 1966.

- (1994) 'Sociale segregatie en functionele samenhang in de regio Amsterdam.' pp. 45-74 in Bovenberg, Korthals Altes and Van der Valk (eds.) op. cit.

Wallagh, G. (1994) Oog voor het onzichbare: 50 jaar struktuurplanning in Amsterdam 1955-2005. Assen: Van Gorcum.

Wissink, B. (n.d.) 'Advocacy and Counsel in Dutch Regional Planning'. Working paper. Amsterdam: AME - University of Amsterdam.

2. Amsterdam in the Red Queen's Country

Interorganisational Co-ordination and EU-Local Interaction in Spatial Planning and Policy

Introduction

The focus of this study combines two perspectives, one procedural and the other substantive. Its procedural orientation is based on an interest in interorganisational co-ordination (IOC); that is: how do organisations interact so as to concert their decisions and actions? This question is relevant for planners and planning, because a good deal of planning is actually anticipative co-ordination (Alexander 1993).

In the 'structuration theory of IOC' I suggested that interaction between organisations is the result of interdependence. IOC ensues through the development (creation or transformation) of co-ordination structures, which range from boundary spanners and liaisons at the micro-scale, through lead agencies and unitary organisations at the meso-scale, to interorganisational networks, mandated frameworks and markets at the meta-scale (Alexander 1995). The European Union (EU) itself is an example of institutionalised IOC resulting from member countries' efforts to deal with their mutual interdependence. Today's EU institutions are the product of its history and the development of a complex system of IOC structures which is still in progress.

Spatial planning and policy is an area of particular interest from the aspect of IOC. This is because it is a 'facet', as distinct from a sector (to use Dutch terminology-see e.g. Ter Heide 1992: 144; Siraa, Van der Valk and Wissink 1995: 60-61, 96). This means that, rather than being defined as a particular sector (with its implications of disciplinary and professional specialisation and functional jurisdiction), spatial planning and policy is a domain that links between sectors, or in which several relevant sectors overlap. As a facet, then, spatial planning and policy makes demands of intersectoral co-ordination beyond the IOC required in any single sectoral policy or planning arena.

EU spatial policy and planning illustrate this proposition very well. Though there is still no explicit EU spatial policy, the elements of what can clearly be recognised as an emerging spatial policy reflect the spatial dimension of what were originally purely sectoral concerns. One of these elements is an emerging regional development policy. This is the result of the 'social cohesion' aspect of EU economic development, and its manifestation in the spatial allocation of EU structural funds in general and the European Regional Development Fund (ERDF) in particular. Another is transnational cross-border planning initiatives, the product of member states' interdependence and the need to co-ordinate their respective spatial development policies in these areas. Others are sectoral regulatory policies with spatial impacts, such as Single European Market (SEM) and environmental regulations, community distributional programs with spatial aspects and consequences (e.g. RENAVAL-a program for adaptive renewal of obsolete harbour facilities), and sectoral programs with significant spatial dimensions, such as DGVII's Trans European Networks (TENs) in transportation and telecommunications (Williams 1996).

Where the procedural focus of this study, then, is IOC, its substantive focus is the development of EU spatial planning and policy. The questions which will be explored are: How do organisations interact in the development and implementation of EU spatial policy and plans; what kinds of IOC models are coming into existence in this emerging policy arena, and how are they structuring the formation and operation of EU institutions and the enactment of EU policy?

EU Institutions and EU-Local Interaction

How and why the EU has developed as it did, and what process accounts for its present and emerging institutions, has been the subject of considerable interest. One aspect of this question has been the interaction between national interests and policy and supranational 'European community' interests and institutions. Two alternative hypotheses have been suggested for this process.

One is the 'intergovernmental' model. This suggests that EU policy is mainly the result of the interplay between national interests, and infers that EU institutions play mainly a mediating role, and serve as arenas for the unfolding of policy debate and bargaining between member states. The other is the 'communautarian' model, giving primacy to the supranational role in the development of EU policy and its institutions. This model suggests that EU policy is made in EU institutions, which have acquired a substantive (if not formal) federal authority to act independently of its member states and their interests (Corbey 1993: 11-15).

Testing these hypotheses (and other modifications of these models), Corbey (1993) explains European integration as a form of 'dialectic structuralism'. This suggests a dialectic process alternating more supranational or communautarian periods of accelerating integration with more intergovernmental intervals of slower integration. The factors behind this dialectic posit a 'structuralist' explanation for the interactions producing this process, relating it to relevant social and economic sectors. Member states' awareness of their interdependence and their common interest in reducing their mutual competition in particular sectors generate heightened integration and the evolution of supranational institutions to regulate those policy arenas. In reaction, this accelerated integration is followed by an interval of slower integration, while member states assert their particular interests by pre-empting supranational policy expansion in adjacent policy sectors.

This thesis is well demonstrated with examples from the history of the European community, and is quite persuasive. But it is less applicable to the question raised here: how is EU spatial policy made? and how does this process affect (and is itself affected by) IOC structures in and among relevant EU institutions? This is because the intrinsically 'facet' and nonsectoral nature of this policy domain makes it less amenable to the 'structuralist' and sectorally-based aspect of Corbey's model.

As possible explanations, then, we are back to the two a priori models suggested above:

• Supranational-communautarian. This implies a top-down policymaking process, and some forms of relatively hierarchical IOC structures integrating the interorganisational systems in the spatial policy domain that link EU institutions with national and subnational agencies and organisations.

• Intergovernmental, suggesting that EU spatial policy emerges in a process of bargaining between member states pursuing their particular interests in various sectors. The intergovernmental model implies the primacy of member states' IOC systems, which 'meet' in a variety of EU IOC structures that serve as negotiating arenas.

Both these are ideal types, and are almost certainly radical oversimplifications of reality. Consequently, we may well expect to find a third model that best represents the true situation. This could be a more complex combination involving either or both of the above explanations with (an)other model or models⁴. Having identified a set of possible hypotheses for testing, let us define the subject of this study.

The substantive topic of this inquiry is the development of EU spatial policy. The subject of our study, then, should be the makers of EU spatial policies. The present problem, however, is that if we know too little about how EU spatial policy is made, we do not know a priori who its makers are and what their respective roles might be. Thus, the potential universe of this study is the whole European spatial planning and policy community: all the potential actors in the European spatial policymaking and planning process. This encompasses EU agencies and institutions, relevant European NGOs, all member states' national and the involved subnational institutions, governments, public agencies, NGOs, interests, organisations and even relevant groups and individuals.

One can imagine this (using a metaphor common among American political scientists) as a 'marble cake'. Its vertical dimension displays socio-political levels of organisation: local/sectoral, regional, national, supranational-European or international. Its horizontal dimensions represent various other kinds of differentiation: by location, cultural and/or political affiliation, and by socio-economic sector. The 'marble cake' image is invoked, rather than the alternative 'layer cake', to suggest a complex mix rather than simple hierarchical divisions and differentiation between the 'cake's' constituent 'ingredients'.

This whole 'cake' is indigestible, in terms of the present study. So our subject is a small, kind of cone-shaped 'slice' of this cake, taking the community (gemeente, or local authority) of Amsterdam as its (arbitrary) point of origin, and working 'upwards' and 'outwards' to trace its relationships and interactions with other parts of and actors in the EU spatial policy community. In particular, our focus is on the interaction between local actors and EU institutions in the process of developing and implementing EU spatial plans and policies. Specific questions to be addressed are:

• What kinds of IOC structures have evolved to handle these interactions? What are their specific tasks and functions, and how effective are they in their fulfilment?

• How does EU spatial policy (and specifically, spatial policy affecting local governments) emerge in the interorganisational network we will have identified, and how is it implemented? Some of the possible alternative hypotheses answering this question have already been suggested above.

A priori we can envisage three types of local-EU interaction.

1. Marketing involves attempts by national and subnational governments and agencies to acquire EU institutions and attract economic activities from outside their borders. It includes the following two elements which are often handled differently:

• The location or relocation of EU institutions and agency offices; this is often undertaken in the same fashion as other EU policy lobbying activities;

• General marketing of the country, region, or locality as a site for locating or relocating European-based economic activities. National, regional and local agencies involved in this activity usually enter the relevant market just like any other landowner seeking potential developer or investor partners.

2. In resource acquisition, national, regional, and local government agencies and other organisations interact with EU institutions to obtain subsidies and funding. The incentive of resource acquisition turns these actors into agents of the EU in implementing its policies and programs. In idealised or simplified form this would be identical to the hierarchical relations between subnational units and national institutions, simply transposed from the national to the supranational level. If things were that simple, this domain of interaction might be trivial, but in fact the reality of Europeanisation adds several dimensions of complexity which add interest.

Among questions it raises are: What, in the process of resource acquisition, is now the role of the national institutions that were, in the 'national' model as contrasted with the European one, the 'end points' or destinations of these interactions? And, as the new 'patrons', is the interaction of EU agencies with their national, regional and local applicants for funds and subsidies and subsequently their project or program implementors the same as that of their national counterparts was, and if it is not, how is it different and why?

3. Developing and influencing EU policy demands participation in the decision-making process that results in EU policies as expressed in EU regulations, directives, decisions, and programs. This means interacting with the relevant other institutions and agencies which are responsible for policies that may have impacts affecting the respective unit, or in which it has an interest. At first sight, for local authorities, such policies may not be apparent, but looking at the elements of EU spatial policy identified above, some candidates become obvious.

It is also important to realise that these three activities are highly interdependent. For example, some policies (such as eligibility criteria for subsidies and participation in EU programs) can affect later marketing and resource acquisition efforts. Consequently, anticipatory involvement in developing and influencing policy may be essential for successful marketing and resource acquisition too. Some of the examples below illustrate this very well.

In most previous work, both the 'supranational-communautarian' and the 'intergovernmental' models of EU policy development have assumed away the roles and interests of most subnational actors. But a wealth of previous experience and observation suggests that these are gross oversimplifications. The role of subnational and sectoral interests in EU policymaking has been recognised and institutionalised in various ways. These include regional representation in the Council of Regions (CoR), representation of subnational sectoral interests (such as employers' associations and trade unions) in ECOSOC (the Economic and Social Committee of the EU), and the representation of a variety of subnational interests, such as environmental NGOs, on subcommissions of the CoR.

But the roles of subnational units of government (regional governments and local authorities) and their interactions with the EU and other actors in the interorganisational networks that develop and influence EU policy have not been systematically explored. This case, then, is a kind of pilot project for which this issue is of particular interest.

'European' activities in Amsterdam

These were part of a larger package, and the outcome of what was seen in retrospect as a totally misconceived and badly managed lobbying effort was that the Netherlands received none of its 'wish list'. This included the chair of the ERBD for its retiring Prime Minister Lubbers, and Dutch locations for the ERBD, EUROFED, and two other smaller EU agencies. Finally, the Netherlands' only consolation prize was locating the EUROPOL headquarters in the Hague (Van Schendelen 1995: 68-70).

One lesson from this experience for Amsterdam, however, together with all the other Dutch actors involved, was that success cannot depend on 'big brother' in The Hague. To influence EU policy and obtain EU resources, you have to 'do it yourself'. This lesson was already being internalised in 1989, when a new Wethouder (alderman or councillor in the municipal executive) who assumed the European/international relations portfolio in the municipal cabinet initiated a formal evaluation of Amsterdam's municipal government European related interactions and activities. The resulting report (Gemeente Amsterdam 1990) proposed a new program which included the Bureau Eurolink in the cabinet department of Economic Affairs to stimulate and co-ordinate Amsterdam-EU relations, and a city underwritten 'Risk and Guarantee' fund to assist potential initiators of European projects.

Bureau Eurolink was to serve as an information exchange node between EU sources and the other agencies of the municipal government. In the report this was described as a two-way information flow. In one direction Bureau Eurolink would generate an information supply from Brussels; this was done by engaging a Brussels based consultant who is the source for all the relevant information and transmits it regularly to Bureau Eurolink in Amsterdam. The other direction was called the 'Amsterdam-EC Project-stream'. Interacting with departments of the municipal government, Bureau Eurolink would search out and identify promising EU-oriented proposals, and provide facilitating services to potential project initiators.

Today Bureau Eurolink operates as a small co-ordinating unit for EU-related activities in the Amsterdam municipal government. It has several tasks. One is to facilitate the EU-related activities and projects which are initiated and undertaken by the sectoral divisions and units on their own. Another is to explore the potential for EU-related interactions in the municipal government's various departments. Finally, and perhaps most important, Bureau Eurolink serves as a conduit and diffusion mechanism for intelligence on EU policy and program related development, which is provided by its Brussels consultants.

In carrying out its first task Bureau Eurolink has become a well-publicised destination for inquiries from other departments, and assists them in advancing their European projects. In performing its second function it executes periodic 'sector audits' of municipal departments, to identify promising areas of EU-related activity. The main medium of Bureau Eurolink's third function, information transmission, is its 'Eurolink Memos', which diffuse information supplied by Amsterdam's Brussels consultants. The Eurolink Memo, which appears monthly, is circulated among all the municipality's departments, and is also distributed to some outside agencies, including (in an English version) other European cities.

Besides the preparation and distribution of these memos, Bureau Eurolink's other interactions with the municipal government's other units are informal. Nevertheless, Bureau Eurolink's staff seems to be well networked in the municipal government, and has acquired a good overview of all the municipality's EU-related activity, as evidenced in its periodic reports. These provide more than anecdotal evidence of the kinds and range of EU-related activities in the municipal government, and confirm intuitive impressions that Amsterdam-EU interaction is almost exclusively sectoral. These activities are not formally co-ordinated in any significant way, though, as we shall see below, there is some informal co-ordination.

Eurolink's (1996) summary of 1995 EU-related activities provides evidence of extensive local government-EU interaction, in a range of undertakings that cover the three types of interaction presented above: marketing, resource acquisition, and developing and affecting EU policy. Marketing goes on all the time; it is done entirely independently of other municipal agencies by the city's Real Estate and Land Policy Division. As an independent activity that is only weakly related to the focus of this paper's concerns, we will drop marketing from further discussion.

Resource acquisition activities consist of implementing EU programs and projects, through applications for subsidies and grants. These take three forms. One form is direct and unilateral initiatives by Amsterdam government departments to carry out EU programs and projects in Amsterdam. Two examples (of 17 projects in 1995) are:

• Physical rehabilitation of the Bijlmermeer (neighbourhood) combined with socio-economic renewal, funded (4.65m ECU 1995 / ca. 2002) by the EU URBAN program-initiated by the municipal Development Department.

• A photovoltaic energy project in Nieuw Sloten: under the EU THERMIE program, carried out by Amsterdam's Energie Noordwest public utility.

This form of resource acquisition is relatively uninteresting for the purpose of this discussion, because it is essentially the same kind of structured hierarchical interaction that also takes place between local and central government. On their face, these projects are the results of proposals or applications in response to EU program announcements and RFPs. However, they are not uninteresting to local governments, because they make a significant contribution of outside (here EU) resources to the local community.

The second form of resource acquisition is when Amsterdam takes part in multilateral EU aid programs directed at other (EU member states or other countries) recipients. Though these activities are quite extensive, they are less significant, of course, in terms of their rewards to the local community: funding to the municipal agency to support its (outside-directed) activity. Besides the intangible benefits of providing a valuable service and acquiring some useful experience, the only direct tangible reward is some contribution (probably relatively insignificant) to the agency's overhead costs. Two cases (of ten in 1995) are:

• Assistance in reconstruction of the Hrasno quarter in Sarajevo), funded by the EU Commission; municipal Housing Service, in co-operation with Cardiff, UK.

• Creating local government economic development capacity in Kiev; funded by the EU ECOS-OUVERTURE program; Economic Affairs Division jointly with the city of Leipzig.

These activities are more interesting than the first type, because they represent much more than conventional local-central government agency interaction, and usually involve a significant degree of networking, as will be discussed below. Networking itself is the focus of the third form of resource acquisition, which involves acquiring EU support for local capacity-building in the shape of interorganisational networks. This already shades into the other area of local-EU interaction, developing and affecting EU policy (which will be discussed next) but there is a subtle difference.

Here, under resource acquisition, networking projects are identified that have capacity building as their explicit focus (though of course some ultimate policy impact cannot be excluded as a hidden agenda). Two examples (of 15 in 1995):

• Assistance in establishing an environmental organisation in Budapest; issues include development of alternative energy sources, waste management, building technology and public transportation. Under Amsterdam's VBM (environmental bureau) as project leader, in co-operation with Lisbon and Budapest, funded by ECOS and the ENVIROPROM program.

• Setting up an integrated framework for the creation of child-friendly open spaces, in co-operation with Neapolis/Thessaloniki (Greece), Brent and Wrekin (UK); initiated by the ISR bureau in the municipal administration, and funded by the PACTE program.

These last two forms of resource acquisition have in common an increasing networking component, which they share (to an even greater degree) with the third type of local-EU interaction: developing, influencing and affecting EU policy. This is the type of interaction that interests us most. What is significant about it is that, in the course of this study, I have not discovered any direct links between Amsterdam local government units and agencies, and EU agencies and institutions. Nor is there any other evidence of the existence of any such links.

Two traits characterise local government activities in developing or influencing EU policy:

• At the local level, these activities are carried out entirely by sectoral units, without any formal or central co-ordination. That is not to say, however, that they are uncoordinated, as we will see.

• All the linkages between the local community and its agencies, and the EU's institutions, are through networking: the creation of and active participation in multimember special interest networks. These may be quite small (sometimes only two or three members) but they are usually much larger, and quite often are Europe-wide (i.e. they contain a representative sample, if not the whole 'population', of relevant organisations from EU member states).

To learn how this works, let us review three cases.

'Creating' a EU Housing Policy

More recently, he approached Bureau Eurolink to discuss how he could create a more institutionalised and recognised discussion platform. Eurolink assisted him to develop a proposal (formally by the Amsterdam Circle) which he presented to the Social Committee of EUROCITIES, a EU recognised network representing a broad array of European local governments. This group proved to be more than willing to make the Amsterdam Circle's proposals part of its policy agenda, and suggested absorbing the 'Amsterdam Circle' and making it a EUROCITIES subsidiary. In the Economic Co-operation Committee of EUROCITIES the Amsterdam Circle's proposal received the same reaction. An intercommittee EUROCITIES initiative proposing an EU housing policy agenda is currently under discussion, which, if it takes its designated course, may result in achieving the Amsterdam Housing Services' Director's original aim: the EU's developing and adopting a policy for EU intervention in the housing arena.

EU Transportation Policy - High Speed Rail (HSR) Networks

High speed rail networks preceded EU transportation-related initiatives, with France's TGV line between Paris and Lyons. But they became an important element of EU transportation policy, with DGVII's development of the Trans European Networks (TENs), that comprise transportation and telecommunications infrastructure. Like many other EU sectoral policy domains, transportation is the subject of intensive interest by and interaction with special interest networks.

A traffic/transportation planner in Amsterdam's Spatial Planning Service is the informal EU-related activities co-ordinator in his unit. He represents Amsterdam on the Transportation Working group of the EUROCITIES network, involving several meetings each year and a good deal of 'homework'. The HSR issue emerged in EUROCITIES as an overspill of a French domestic debate. This debate was in reaction to the perceived overcentrality and power of Paris, and involved other French cities' attempts to increase their own centrality through HSR rail linkages.

In particular Lyon (already linked with Paris by TGV) wanted to become a subnode in a more extended system, and promoted the idea of a southern European HSR network. Lyon recruited other cities to join its initiative; initially these were mostly southern and Mediterranean cities, but later other northern cities, including Rotterdam and several German cities, joined it.

At present the Working Group is studying the impacts of HSR stations, and addressing the nature of linkages between HSR systems and airports, both issues with very concrete relevance for Amsterdam. They are of high current interest, since their implications include the location and number of prospective HSR stations. One positive result of Amsterdam's involvement in this Working Group is the suggestion of some basic analysis that two HSR stations are a real possibility for some cities (such as Amsterdam) and may even have some decided benefits. Another implication relates to future HSR routes (also identified as TENs) linking Amsterdam to other European cities. The one firm project is the TEN connecting Amsterdam southward (with a station at Schiphol) to Brussels, Paris and beyond. Current discussions suggest a future HSR link eastward to Germany; these also have a bearing on Amsterdam's future development plans and the priorities for HSR station locations.

Major European Airport network to influence EU policy

The Airport Regions Conference (ARC) was founded in 1994; among its initiators was a member of the agency in the Province of North Holland which is concerned with Schiphol airport. An active participant representing Amsterdam is the (now ex) Director of the Economic Affairs Division, who was (and continues to be) working on development of the Schiphol area. ARC's members are local governments and regional authorities with major airports in or adjoining their jurisdictions.

A formal organisation, ARC receives funding from the EU through the RECITE program, and is governed by an Executive Board currently chaired by the Head of the Europe Office in West Sussex County Council's Chief Executive Department. His agency also provides ARC with necessary staff services, which are critical since ARC's member representatives all undertake their ARC-related activities in addition to their regular full-time jobs.

ARC is organised into three Working Groups. One, Airport-related Business, is concerned mainly with information exchange between members, focusing on airports' economic development potential. This includes data , such as employment multipliers, which can be useful to members in making their airports' case at home, and the development of common management practices that will improve members' economic development performance. Amsterdam's interest in this Group is largely altruistic: it sees its role, as one of the European leaders in airport economic development, as sharing its knowledge and experience with members of ARC who are less advanced in this field.

The second Working Group, Accessibility, is similar to the first, but in a different area of concern. Its domain is the infrastructure-roads, public transport, and mass transit-providing access to and from the airport. In this area, too, Amsterdam's Schiphol Airport is a leader, so Amsterdam's role here too is also largely as a source of information and best management practices.

The third Working Group is Pollution and Noise. While the obvious agenda of this group is also information exchange and mutual learning, its long term purpose is to influence EU policy regulating airport-related environmental impacts. Such policy may include setting uniform maximum noise standards, and requiring appropriate operating procedures to minimise noise pollution.

This Working Group initiated comparative research on members' local measures to regulate and control noise impacts which provided the information base for the development of common policy. At its third annual conference in Helsinki, ARC issued a statement recommending minimisation of night flights because of their disproportionate noise impacts. But this is a very political issue (with quite tangible economic consequences) and the operational interpretation of 'minimise' for different flight control regimes is still unresolved. Meanwhile, their information exchange on this and other issues helps ARC members in local debates on regulatory policy with their own national governments.

Setting uniform environmental standards is a hotly debated issue, in the expectation of ultimately participating in the development of EU environmental regulations. But this is also a contentious problem, with divergent interests among ARC's members. Those representing airports in states with high environmental standards (such as the Netherlands) plead for low maximum acceptable pollution levels. Though their case is made with environmental health arguments, their real motivation is that this would improve their airports' competitive position vis a vis other airports in less demanding contexts (e.g. Amsterdam vs. Athens).

The counterargument comes from members with low environmental standards and often poor enforcement. These, mostly from the less developed European peripheries, claim that higher standards would be incompatible with their countries' environmental control regimes, and would impose economic costs which are disproportionate to the arguably marginal health, longevity and quality-of-life benefits. ARC's resolution of this issue, its members hope, will become a generally acceptable prototype for EU environmental policy.

These cases are only some examples of the widespread involvement of Amsterdam local government personnel in the EU policymaking process. In Eurolink's 1995 Activity Summary 12 such networking activities are reported. They include participation in the Union of European Capitals (UCUE), chairing the European Cities on Drug Policy (ECDP), membership of the information science network EURICOM, participating in the TELECITIES network, and working in Eur-HOPE, a European network of health organisations for the poor and excluded.

IOC Structures for European Activities in Municipal Government

Formal co-ordination of all its EU-related activities in the Amsterdam municipal administration is minimal, limited to information exchange with and through Bureau Eurolink. But there are some arenas for formal intersectoral co-ordination in several 'facet' areas. These are interdepartmental Steering Groups in various policy domains which meet regularly. One example is the Traffic and Transportation Steering Group, which includes representatives from other agencies with an interest in this policy 'facet', such as Environmental Policy. In these Steering Groups EU-related issues, projects and activities are raised and discussed, and thus co-ordinated through information exchange and mutual adjustment by the relevant participants.

What exists here is a model of IOC which is quite institutionalised, in the sense of the relative roles and importance of formal and informal co-ordination. But it is quite loose, in terms of the kinds of IOC structures linking the organisation's units. The Bureau Eurolink is a central co-ordinating unit, but its co-ordination tasks are limited. Of the four kinds of co-ordination that it could do-information exchange and co-ordination by mutual adjustment, task-related operational co-ordination, managerial and administrative co-ordination, and anticipatory co-ordination (strategic policymaking , planning and resource allocation)-it has been charged with and performs only the first.

Intersectoral interorganisational groups (the interdepartmental Steering Groups) offer additional arenas for information exchange, and perhaps some informal operational co-ordination between units engaged in interdependent or parallel EU-related activities. Project-based interdepartmental or interunit Working Groups, are also usually set up as necessary for any EU-related projects (as for other projects in the Amsterdam municipal administration) that reach the implementation stage. There is certainly no anticipatory co-ordination of EU-related activities, in the sense of setting overall goals, determining priorities and allocating resources among organisational subunits of the municipal administration for EU-related activities and projects. With a few exceptions (initiated by Bureau Eurolink), almost all the initiatives for local government-EU interactions are sectoral and reflect the sectoral departments' respective interests. This is not necessarily bad: it may be quite an efficient form of decentralisation. And, in terms of the relative costs and benefits of alternative models of IOC, this model may be quite effective.

Interestingly, there are several alternative models. One was proposed by a consultant, who was called in to evaluate the effectiveness of the municipality's international and European efforts and to recommend any improvements he thought necessary. He proposed creating a much larger co-ordinating unit for all international activities, a Bureau Buitenland (Foreign Office) attached directly to the Mayor's Cabinet and the Wethouder responsible for International Contacts.

This co-ordinating unit would carry out the full gamut of co-ordination tasks. It would provide operational co-ordination through an institutionalised consultation process, and information exchange, including the absorption of Bureau Eurolink and taking over its tasks. It would become a one-stop contact point in the municipal administration for outside actors and delegations, and apparently it would perform some policy and anticipatory co-ordination through its control of budget allocations for international activities and travel (Spits 1996).

Clearly this is a much 'tighter' model than Amsterdam's present IOC system, but it raises several questions. One is the adequacy of its proposed resources to its assigned tasks. This is a common cause of failure of IOC structures, especially interorganisational groups (Alexander 1995: 141-2, 175-7). Another question is how well this tightly linked IOC structure would fit into Amsterdam's traditionally looser organisational culture.

Another alternative model is presented in a counterproposal by Eurolink staff (Eurolink n.d). Recognising some shortcomings in current practice, it addresses them in a very different way. The recommended IOC structure includes individuals designated as co-ordinators of EU-related and international activities in their organisational units, who would meet periodically as a formal interorganisational group for International Contacts Policy: the Beleidstaf Internationale Contacten. This group would be staffed by a full-time co-ordinator (the 'International Desk') in the municipality's central administration, linked to the International Contacts portfolio in the City Council.

This is a much looser model, envisaging largely the same basically consultative mode as is practised today. The main changes it proposes are some normalisation of the departmental co-ordinator's role, and a bit more centralisation of the intraorganisational network configuration. While this IOC system seems very compatible with Amsterdam's consultative organisational culture, perhaps it fits too well if more radical changes in organisational performance are desired.

Rotterdam offers yet another model, which separates international relations from European affairs. The former are managed by a co-ordinating unit (with a staff of 12): the International Affairs section in the city's General Administration (Algemene Bestuurszaken) Directorate. EU-related activities, on the other hand, are run out of the Rotterdam Development Agency (Ontwikkelingsbedrijf Rotterdam) which houses a co-ordinating unit that oversees EU-related project development and interactions with EU institutions. Nevertheless it appears that Rotterdam is more like Amsterdam than it seems, judging from its apparently greater centralisation compared to Amsterdam today. In Rotterdam, too, local government-EU interactions related to sectoral EU funding and policy are left to the relatively undirected initiative of the sectoral agencies (Lange 1993; Spits 1996).

In both these cities, a great deal of local-EU interaction takes the form of networking, which is channelled by policy domain or sector, special interest orientation, or focused by the particular issue. This raises the question of whether any benefit is to be gained if these activities are subjected to more formal operational and management-administrative co-ordination, or if they were more integrated and prioritised by some central policy co-ordination unit. Undoubtedly, the response to this question depends on particular circumstances.

Discussion and Conclusions

Although answering the last question posed is beyond the scope of this paper, there is much to be learned from our exploration of Amsterdam municipality-EU interactions through the lens of IOC. This can be discussed under several headings.

Community-level internal co-ordination of EU-related activities

Amsterdam has a 'loose' IOC system: a small co-ordinating unit limited to information exchange and informal co-ordination by mutual adjustment. This is also informally linked to some intersectoral co-ordination that takes place through several interorganisational groups, steering committees (stuurgroepen), which integrate the work of several departments in related sectors. These may also effect some relatively informal operational and management co-ordination of related activities.

Alternative, 'tighter' models exist and have been proposed. However, all efforts at co-ordinating EU-related activities in local municipal administrations have to take into account the salient characteristics of the activities they are co-ordinating. These are that they are essentially sectoral in their orientation, and that they consist largely of networking in a wide range of very diverse and separate interorganisational networks. Tighter and more centralised co-ordination, like Rotterdam's, does not change the intrinsic nature of these activities.

It is very likely that the cases we have observed are quite typical of EU-related activities in similar European municipal administrations. They suggest that by their very nature such activities, to be effective, demand an organisational environment that recognises diversity (e.g. between separate sectors, their different disciplinary/professional orientations and their various interorganisational networks) and rewards local initiatives.

More tightly coupled co-ordination systems may provide some efficiency gains, at the level of anticipatory co-ordination through prioritising, monitoring and controlling the resources allocated to these activities, and at the operational level by co-ordinating parallel or interdependent activities, though the latter are likely to be few. But such systems also have their costs, among which is their possible inhibiting effect on these very dispersed activities and local initiatives which are to be encouraged. These trade-offs must be considered by each local government for its particular case.

Local-EU interaction in EU spatial policymaking

The more interesting question, from our perspective, is: What are the implications of EU-local interaction, as observed in Amsterdam, on the making and implementation of EU spatial policy? Obviously any answers offered here to this question are somewhat speculative, based as they are on a single case. But to some extent they are supported by and confirm other studies and evidence (e.g. Van Schendelen 1993, 1995; Williams 1996).

It is clear that the possible model of European integration by the evolution of supranational institutions, and its implied process of simple hierarchical 'top-down' development and implementation of EU policy, are poor descriptions of reality. This model is contradicted on its face by the intensive, complex and multifaceted local-EU interactions we have observed.

This is obvious, though it is qualified by the observation of a limited hierarchical 'supranational' policy implementation process as part of the much larger and more complex overall set of local-EU interactions. This is the process of implementing EU projects and programs through the conventional sequence of program development and announcement, distribution of Requests for Proposals, processing and evaluating local government's responses and funding requests-the local program/project application and funding process. If the distribution between various kinds of EU-related activities observed in Amsterdam is typical, the kind of resource acquisition activities which are the local counterpart of this type of EU policy and program implementation, are only a relatively small part of the whole gamut of local-EU interactions.

Incidentally, the apparently limited role of national government agencies, or their complete absence from these hierarchical types of interactions is noteworthy. In my view, this provides clear evidence of the emergence of some degree of supranationalism, contradicting the intergovernmental model of EU governance and policy development. This has some interesting longer-term implications which will be discussed later.

In terms of the decision making process accountable for the development of EU spatial policy, clearly both the simple supranational-communautarian and intergovernmental models are totally inadequate. Not surprisingly, each of them fails to describe or explain the much more complex reality. Another simple alternative, a 'bottom-up' process of EU policy development, is also obviously incorrect. This model implies face-to-face interaction between EU agencies and institutions, and subnational and local agencies, organisations, and interests, and the direct involvement of the latter in developing or influencing EU policy. This simple model is wrong, too, because it is evident that no such direct EU-local interaction ever takes place.

If all these explanations are incorrect, what, then, does happen? The salient feature of the model of EU policy development which we have observed is networking. Almost all the interactions of subnational and local agencies with EU institutions are mediated through interorganisational networks, many of which have come into existence for this very purpose. These networks are, at their smallest and simplest, multinational, and at their largest, Europe-wide. They take various forms: policy arena-related, common interest based, or special issue focused. But there is hardly any EU-local policy-related interaction except through such an interorganisational network.

Even if they are born as informal associations (as some of them are) these interorganisational networks almost always become quite institutionalised. Often such a network has its own central co-ordinating unit in the form of a permanent director and staff; sometimes these are supplemented by consultant services, and sometimes they are staffed by member organisations. Like all such networks, they are governed by interorganisational groups: a Board, Commission, Council or Directorate made up of (sometimes all, sometimes selected) representatives of member organisations. The network's operational costs are borne by its member organisations, sometime with some outside (e.g. special interest foundation) support. Not infrequently the EU itself, through appropriate programs, provides some or all the network's funding.

This case suggests (and other observers agree) that interorganisational networks are what mediates local-EU interaction in stimulating, developing, or influencing EU policy in areas of interest to local government. This has some significant implications for how interested localities must view and manage these interactions. Influencing EU policy is a very indirect and complex process. It demands long lead-times and significant front-end investments in time, personnel and resources for rather uncertain payoffs (in the form of favourable regulative policies or actual resource allocations: grants or subsidies) in a relatively distant future.

Networking and its implications for EU integration

What are the implications of these findings for our views of European integration? One implication must be that the proliferation of interorganisational networks as mediators between the EU and its policy environment is evidence of the gradual erosion of intergovernmentalism, if the intergovernmental model of EU decision making was ever true.

Member states' governments' and agencies' roles in developing EU policies seem less frequently to be taking the form of direct interaction with each other and with EU agencies in the various existing EU policymaking arenas. Instead, they appear more often themselves to be mediated through the interorganisational networks of which they are members. The degree to which national governments' role is being supplemented (and perhaps, sometimes, even supplanted) by interorganisational networks in which they have no part enhances this trend.

In our view from the bottom of the 'marble cake' of intergovernmental (here including EU) relations, it looks as if interorganisational networks that do not include national government agencies: associations of substate and local-level agencies and organisations representing functional, sectoral, interest, or issue-based constituencies, are playing an increasingly important role in the EU policymaking process. The inference of declining intergovernmentalism is unavoidable.

Another implication concerns the probable growth of this model of European integration, which is based on the symbiosis between the kind of networking observed and a limited supranationalism. European networking itself is the result of a coincidence of interests between communautarian EU institutions and subnational constituencies. The most tangible evidence for this is the frequent encouragement of network formation (formal and informal) by EU agencies, and their widespread (if not universal) support with EU funds.

If what we are seeing is a self-reinforcing system, the EU policymaking process will increasingly come to resemble the one observed in federal polities such as the US, with EU institutions growing to fill the role of a European federal government. Like US States, European national governments will have their formal institutionalised participation (through their representatives in EU institutions) in making EU policy, but (again like the States in the USA, which are-Europeans believe it or not!-also sovereign governments) their substantive role will be residual. The growing role of European networking in making EU policy is what may integrate Europe into an essentially federal state, and make national governments increasingly redundant.

References

Alexander, E.R. (1993) 'Interorganisational coordination: Theory and practice', Journal of Planning Literature 7, 4: 328-43.

Alexander, E.R. (1995) How Organisations Act Together: Interorganisational Co-ordination in Theory and Practice. New York/Montreaux: Gordon & Breach.

Barrett, S. and C. Fudge (eds.) (1981) Policy and Action: Essays in the Implementation of Public Policy. London: Methuen.

Corbey, D. (1993) Stilstand is Vorruitgang: De Dialectiek van het Europese Integratieproces. Assen: Van Gorcum.

Economist (1997) May 31st -June 6th Survey European Union: The end of federalism.

Eurolink (n.d.) 'De internationale meerwaarde georganiseerd', Internal memo.

Eurolink (1993) Eurolink Amsterdam: 'Op Zoek Naar Partnership'. Amsterdam: Eurolink - Economische Zaken.

Eurolink (1996) Opsomming van Amsterdamse EU-Projecten/Netwerken/Activiteiten in 1995. Amsterdam: Eurolink - Economische Zaken.

Faludi, A. and A. van der Valk (1994) Rule and Order: Dutch Planning Doctrine in the Twentieth Century. Dordrecht: Kluwer.

Gemeente Amsterdam (1990) Nota Amsterdam in de Europese Gemeenschap (EG). Amsterdam: Afdeling Economische Zaken.

Heide, H. ter (1992) 'Diagonal planning: Potentials and problems', Planning Theory 7/8: 135-54.

Holland, M. (1994) European Integration: From Community to Union. London: Pinter.

Jeffrey, C. (ed.) (1993) Federation, Unification and European Integration. London: Cass.

Lange, C. de (1993) 'Lage overheden lobbyen voor EG susidies'. pp. 219-50 in M.P.C.M. van Schendelen (ed.) Nederlandse Lobbys in Europa. 's-Gravenhage: SDU Uitgeverij.

March, J.C., and H.A. Simon with H. Guetzow (1958) Organisations. New York: Wiley.

Mazmanian, D.A. and P.A. Sabatier (eds.) (1981) Effective Policy Implementation. Lexington, MA: D.C. Heath.

Pressman, N. and A. Wildavsky (1973) Implementation. Berkeley: University of California Press.

Pinder, J. (1991) European Community: Building of a Union. Oxford: Oxford University Press.

Schendelen, M.C.P.M. van (ed.) (1993) Nederlandse Lobbys in Europa. 's-Gravenhage: SDU Uitgeverij.

Schendelen, M.C.P.M. van (1995) Gelijkhebben of Winnen? Nederlandse Belangenbehartsigen in de EU. Amsterdam: University of Amsterdam Press.

Siraa, H.T., A.J. van der Valk and W.L. Wissink (1995) Met het Oog op de Omgeving: Een Geschiedenes van de Zorg voor de Kwaliteit van de Lefvormgeving - het Ministerie van VROM (1965-1995). Den Haag: SDU Uitgevers.

Spits, M. (1996) Van Interne naar internationale samenwerking - Gemeente Amsterdam. Exelcom Consultants Public Relations and Public Affairs: Amsterdam.

Welsh, M. (1996) Europe United: The European Union and the Retreat from Federalism. New York: St. Martins Press.

Williams, R.H. (1996) European Union Spatial Policy and Planning. London: Chapman.

Editor: Carina Mulié-Velgersdijk
Last update: