The Dutch ‘Gateway to Europe’ spatial policy narrative, 1980–2020: a systematic review

ABSTRACT Like other countries with large ports, the Netherlands developed a policy narrative to acquire a key position in global value chains starting in the 1980s, through the spatial development of its hinterland logistics complex. The negative environmental effects of logistics, such as landscape transformation and congestion, have increasingly come to be seen as spatial policy problems. The literature on policy narratives emphasizes the importance of balanced trade-offs and learning from alternative views. In this paper, we discuss why the ‘Gateway to Europe’ narrative has remained in place. This paper systematically reviews spatial planning documents, advisory reports and academic papers between 1980 and 2020 to develop a chronology of logistics planning concepts pertaining to economic and technological milestones. It also maps policy influences, aiming to identify underlying causal policy theories on logistics development and its spatial-environmental effects. We determine that critical reports have been structurally ignored, challenges have been outsourced and advocacy coalitions have been unbalanced, increasing path dependency and risking a spatial-economic lock-in. Looking at the ‘Gateway to Europe’, we point to pitfalls in the policy narrative and the policy-learning process, enabling policymakers to avoid them in the future.


Introduction
Since the 1980s, European countries have strategically positioned themselves in the emerging trade paradigms of global supply chains, global value chains and the free flow of capital, people and goods. This positioning process has entailed the elaboration of policy narratives and high-impact spatial planning decisions concerning transport infrastructure and adjacent logistics area developments, together forming the logistics complex. While the European Union (EU) has promoted transnational corridors to enable 'seamless flows', 1 the Dutch have attempted to become a distributieland ('distribution country', formally translated as 'Gateway to Europe'). This policy narrativecreated by the government, economic interest groups and state-owned companies such as the Port of Rotterdammay be considered to be neoliberal, aimed at eliminating companies' transaction costs. It may also be viewed as neo-mercantilist, aimed at enhancing the competitiveness of the Dutch trade and logistics sector ( Figure 1). 2 Large logistics complexes with rising spatial footprints have been developed near Antwerp, Hamburg, Los Angeles and across Central and Eastern Europe over the same time period with support from similar policy narratives. 3 In the Netherlands, the rising number and size of distribution centresresulting in the so-called verdozing (boxification) of the Dutch landscapeis an emerging hot topic in recent spatial planning debates. 4 This phenomenon, often referred to as 'logistics sprawl' in the literature, 5 may be more than just incompatible with established policy goals, such as net-zero emissions and the circular economy. 6 It may also seriously compromise the quality of the Dutch living environment via road congestion, heightened emissions and landscape transformation. 7 Recent research suggests that knowledge-intensive sectors of the economy require a favourable living environment to attract and retain talent. 8 In its most recent planning strategy, 9 the Dutch government established this environmental favourability as a national policy goaland logistics sprawl as a threatalongside the further development of logistics. This contradiction and the handling of various trade-offs in Dutch national policy constitute the main focus of this paper.
Despite concerns over its negative effects, the Gateway to Europe narrative has held strong for over 40 years. This raises questions regarding the assumptions on which this policy narrative was based and how these have changed in light of evolving empirical evidence. The literature on policy narratives emphasizes the importance of balanced trade-offs 10 and learning from alternative views. 11 It also highlights the need to understand the evolving causal policy theories present in  Environment VROM, 1988), translated by the authors. Right: Investments since 1990 (author's illustration, based on online sources, including MIRT project books). 3 Flämig and Hesse, "Placing Dryports"; Frejlachová et al.,Steel Cities;De Lara in Hall and Hesse,Cities Regions and Flows. 4 CRa, Rademacher & De Vries and Stec Groep, (X)XL-Verdozing. 5 Krzysztofik et al., "Beyond 'Logistics Sprawl' and 'Logistics Anti-Sprawl'"; Strale, "Logistics Sprawl in the Brussels Metropolitan Area." 6 Van Buren et al., "Towards a Circular Economy"; I&W and EZK, A Circular Economy in the Netherlands by 2050; BZK, "Nationale Omgevingsvisie-NOVI"; Fichter, "E-Commerce." 7 Aljohani and Thompson, "Impacts of Logistics Sprawl on the Urban Environment and Logistics"; Heitz, Dablanc and Tavasszy, "Logistics Sprawl in Monocentric and Polycentric Metropolitan Areas." 8 Vereniging Deltametropool, Blind Spot-Metropolitan Landscape in the Global Battle for Talent; Rli, "Mainports Voorbij." 9 BZK, "Nationale Omgevingsvisie-NOVI." 10 Surel, "The Role of Cognitive and Normative Frames in Policy-Making." such narratives. 12 Therefore, in this analysis, we aim to identify the weaknesses of the Gateway to Europe policy narrative dating back to its emergence in the 1980s.
The following section draws on three elements from the literature to define the structure of our systematic review method: exogenous events, influence by advocacy coalitions and independent research, and causal policy statements. We apply this method to carefully selected spatial policy memoranda, policy advisory reports, relevant research and academic papers, from which we extract information on the development of the Dutch logistics complex and its spatial-environmental effects. In the third section, we structure our findings as a timeline of policy concepts and instruments, a table of policy influences, and an overview of the main causal statements in the policy memoranda. In the fourth section, we reflect on our findings through the lens of policy narratives and policy learning.

Key analytical concepts
We intend to show that the Gateway to Europe narrative is not only a public-private policy narrative but also a policy-learning process, the results of which are suboptimal relative to what these concepts entail. Policy narratives have been increasingly theorized since the 1990s. They are broadly understood as causal stories aimed at mobilizing support for a project 13 or 'side-step [ping] opposition from potential losers and avoid[ing] policy deadlocks.' 14 To begin our inquiry into Gateway to Europe, we can look to Peter A. Hall's practical definition of a policy narrative: [T]he terms of political discourse generally have a specific configuration that lends representative legitimacy to some social interests more than others, delineates the accepted boundaries of state action, associates contemporary political developments with particular interpretations of national history, and defines the context in which many issues will be understood. 15 Planning often comes down to persuasive and constitutive storytelling: future-oriented texts that, according to Throgmorton,16 not only pertain to the planner's own ideas but also 'reflect awareness of differing or opposing views.' This storytelling involves rhetorical framing, 17 in which deliberately chosen adjectives, nouns and metaphors are used to achieve the political and societal acceptance of policies and interventions. Spatial narratives, the category to which the Gateway to Europe narrative partially belongs, often include 'framing with images.' 18 For instance, such narratives may highlight the favourable position of a country in the global trade network.
In its most condensed form, a spatial narrative can be a planning concept. The Gateway to Europe narrative contains several such concepts, the most important one being the mainport. Planning concepts combine analytical and empirical explanations of spatial elements with normative statements on spatial policy goals. 19 Some concepts become dominant spatial imaginaries, viewed as true representations of reality. 20 One example consists of contemporary Eurasian trade links, collectively imagined as the Silk Road by combining an oversimplified historical reference with the current Chinese Belt and Road initiative. While many planning concepts and policy metaphors remain stable over the course of several decades, their underlying meanings may change. The concept of a mainport, for example, was introduced in the literature as an element of wider spatialorganizational and logistics networks. Spatial-economic and infrastructural policy narratives reduced its meaning to a physical hub with a confined hinterland. 21 Policy narratives are not exclusively developed by either market or state actors; rather, they typically form through public-private collaborations. For this reason, this paper employs the concept of advocacy coalitions. 22 In the context of hinterland logistics, Raimbault argues that purely technical perspectives on transport flows are insufficient to understand developments in the field: 'Agenc [ies] can lobby governments and align with institutional actors to secure interests or pursue development agendas through networking'. 23 Therefore, behind each narrative, there exists a coalition of actors with shared beliefs and ambitions seeking to coordinate in pursuit of desired outcomes. In the context of the Gateway to Europe narrative, these outcomes include port infrastructure, hinterland connections (e.g. roads, rail networks, waterways) and sites for the development of logistics buildings.
Policy narratives contain policy theories: the causal assumptions underlying a policy, 24 including the assumed effects of policy instruments and interventions. These are sometimes-but often not supported by evidence. The Gateway to Europe narrative entails assumptions regarding the positive economic effects and necessity of infrastructure investments. Such policy theories are often biased, underestimating the costs and overestimating the yields of infrastructural megaprojects. 25 Although policy theories are generally not made explicit in policy documents, they can be reconstructed from causal statements across various sources making a distinction between: problem → policy goal → policy instrument. In this paper, we describe only policy theories that can be traced back to statements in official policy memoranda. In the case of spatial policies pertaining to the Gateway to Europe narrative, we find assertions based on evolving economic conditions, production chains or transport technologies. This brings us to what's often referred to as policy learning. According to Surel,26 two types of events are likely to prompt changes in the analytical and normative underpinnings of policy narratives and the composition of supporting advocacy coalitions: shifts in economic conditions and exogenous shocks to policy subsystems. Clearly, the interpretation of exogenous events by planners is of great importance to our case. Spatial planning is increasingly viewed as a learning process 'concerning collaborative action and future challenges regarding society, economy and natural environment.' 27 The planning discourse around the Gateway to Europe narrative is potentially such a learning process. Over the last 40 years, the empirical basis of spatial planning knowledgeand, more specifically, of the Gateway to Europe narrativeremained rather narrow, as will be shown below. The empirical basis has relied on the observation of a limited number of cases, and there is a normative bias in terms of what planners and decision-makers perceive as 'valid' and 'relevant' knowledge. 28 Therefore, the learning process not only relates to policy theories 20 Sykes & Shaw in Davoudi et al., "Policy and Practice Spatial Imaginaries." 21 Van Duinen, Planning Imagery; Van Duinen, "Mainport and Corridor." 22 Sabatier, "The Advocacy Coalition Framework." 23 Raimbault, "From Regional Planning to Port Regionalization and Urban Logistics," 2. 24 Hoogerwerf, "Reconstructing Policy Theory"; see also, Rodrik, Straight Talk on Trade, 165. 25 Flyvbjerg,Bruzelius,and Rothengatter,Megaprojects and Risk. 26  but also to the normative foundations of policy narratives and their constitutive spatial analyses and planning concepts.
Applying the above considerations to the Gateway to Europe narrative, we may assume that the cognitive and normative frameworks of this narrative have been influenced over the years by interpretations of exogenous shocks to economic and technological conditions, biased readings of empirical insights and the work of advocacy coalitions.

Methodology
Given the complex 40-year history of the Gateway to Europe narrative, a comprehensive account of its main decision-making processes and surrounding sociopolitical debates would be beyond the scope of this paper. We are primarily interested in how the national government has interpreted the Gateway to Europe narrative in spatial policymaking, how the narrative changed over time and how these changes can be explained. Therefore, we focus on formal documents pertaining to spatial policymaking, including national policy memoranda and other sources explicitly linked to the elaboration of such documents (e.g. reports from government advisory bodies, expert hearings, academic articles). We also looked at relevant reports from government advisory bodies that were not requested by the government, as well as research papers that discord from prevailing policy theories. To keep the analysis as transparent and replicable as possible, we use a systematic review to reconstruct the evolution of key policy theories underlying the Dutch logistics complex alongside exogenous shocks, external influence of advocacy coalitions and empirical research. Specifically, we employ the PRISMA 29 method, which requires the explicit documentation of both the selection of sources and the treatment of data. 30 All of the stepsincluding the identification, screening and assessment of sources and the analysis itselfare illustrated in Figure 2. The spatial scope of the selected documents is the Netherlands and other countries in Northwest Europe; the historical scope is from 1980 to the present, capturing the global shift toward neoliberalism and the growth of global supply chains. 31 We ran all sources through a screening process, ensuring that they met explicitly defined criteria before retrieving information from them. The final selection includes six normative policy memoranda, 13 hybrid policy-advice documents and 23 empirical research papers. From the 19 policy and advice documents, we retrieved and elaborated the following information: timing of the document (relating the used policy concepts and instruments to economic and technological milestones on a timeline); declared input by advocacy coalitions and studies (presented in a comparative table to assess the influence on the policy documents); argumentation regarding the development of the logistics complex and its spatial effects (from which the main causal policy theories are distilled).
The 23 selected research papers enabled us to create an overview of the available knowledge at their time of publication, from which we can assess the extent of their use in policy memoranda and advisory reports. The validation of the empirical basis of policy is not the goal of this paper. Since some of the documents were selected with the help of experts, a limited degree of bias may be present in spite of careful triangulation. As we excluded newspaper articles and other such sources, the 28 Balz, "Regional Design." 29 PRISMA, "Preferred Reporting Items for Systematic Reviews and Meta-Analyses." 30

Analytical results
We structure the findings of this systematic review chronologically in the first sub-section to assess which exogenous events (macroeconomic and technological milestones) coincide with policymaking. The next sub-section identifies the advocacy coalitions and empirical research on which policy memoranda and policy advisory reports have been based. The third sub-section distils those policy theories (from the most relevant reviewed documents) with an eye for how the government has viewed the development and spatial effects of the logistics complex.
Historical periods in the development of the 'Gateway to Europe' narrative The timeline (Figure 3) aligns relevant economic and technological events with the introduction of spatial planning concepts and instruments regarding the logistics complex. Major events, represented by larger stars, predate several key logistics policies between 1980 and 2020, some of which are explicitly mentioned in planning memoranda and advisory reports. One key example is the economic crisis of 1981-1982, from which the Dutch economy recovered more slowly than other European countries; 32 this crisis set the scene for a profound spatial-economic policy shift and the Dutch ambition to become a Gateway to Europe. The logistics revolution of the 1970s reorganized supply chains worldwide (see timeline). Alongside the growth in container Figure 2. Flowchart of systematic review, adapted from PRISMA model. 32 Bakker, "Economische Crises Jaren Dertig En Tachtig Vergeleken." traffic in the 1980s, this further stimulated the development of a mainport policy, anticipating the 1992 integration of the EU internal market.
In the hinterland of the port of Rotterdam, emerging e-commerce platforms strongly propelled the development of distribution centres in the 2000s despite the dot.com crisis of the late 1990s. The end of the financial crisis and economies of scale in distribution centres spurred the growth of large 'XXL' distribution centres starting in 2014. Both the dot.com and financial crises were mentioned in policy memoranda. 33 In this context, policy concepts and related instruments (e.g. zoning plans) aimed to expand and establish new distribution centres; these efforts can be understood as reactions to exogenous economic and technological events.
While the Gateway to Europe narrative and the need for an attractive living environment in the modern knowledge economy have remained rather stable narratives over the last four decades, this is not the case for all spatial planning concepts referenced in policy documents. For example, when a government research agency revealed a large increase in new business locations along motorways, 34 the resultant public and political outcry led to the adoption of the concept of snelwegpanorama (motorway panorama). Motorway panoramas were institutionalized in a dedicated policy document two years later, calling for a spatial strategy around motorways to enhance the 'view on the beauty of the Netherlands.' 35 A few years later, the concept was dropped when a new political coalition advocated for a narrower, less interventionist role of the national government in spatial planning. Based on our estimated impact of exogenous events, the evolution of the Gateway to Europe narrative can be divided into three distinct periods, each kick-started by a significant macroeconomic event and marked by important policy and technological events.
In 1980-1991, the Gateway to Europe narrative was being prepared. The main spatial policy of that period was the 1988 Fourth Spatial Planning Memorandum. The relevant political debates of the 1980s focused on the change from stringent spatial-economic planning to public-private development planning as well as the trend toward internationalization: how to prepare the Netherlands for the 1992 EU market integration? Top-down spatial planning and direct investment in infrastructure were seen as tools to enhance national economic performance. A novel spatial-economic vocabulary became a part of national planning, emphasizing the strengthening of the national spatial main structure and its elements, such as hinterland connections.
In 1992-2013, the country's logistics policies became more elaborate, facilitating the convergence of global production chains, container transport and ICT. 36 The 2004 National Spatial Strategy ushered in the decentralization of most spatial planning issuesother than those pertaining to national infrastructureto provincial and local governments and called for private-sector involvement in spatial development. 37 As the logistics complex emerged as a spatial phenomenon, advocates of motorway panoramas failed to achieve effective policies. During the 2008-2013 financial crisis, austerity politics were combined with the deregulation of spatial development guidance, for example pertaining to logistics business estates. The Crisis and Recovery Act (2010), for instance, created temporary shortcuts in planning procedures. These shortcuts are being integrated into the Omgevingswet (Environment and Planning Act), which is expected to take effect in 2023.
In 2014-2020, the Netherlands experienced strong e-commerce growth and economies of scale across its distribution centres, causing friction among policymakers and the public at large. The 2020 National Strategy for Spatial Planning and the Environment, 38 like earlier memoranda, attempted to reconcile the growing spatial footprint of logistics with spatial-environmental considerations. However, due to the aforementioned decentralization, several policy instruments were in the hands of local governments. Today, the national government continues to seek advice regarding its logistics developments; 39 however, at the time of writing, it has yet to decide on a course of action. The 'boxification' of the landscape became a regular item in the debate around 2018, fuelled by civil and political unrest regarding XXL distribution centres, some of which extend across 100,000 square metres. While some of these 'big boxes' house factories or data centres, most have a logistics function.
The influence of policy advice, advocacy coalitions and research in spatial policymaking Figure 4 summarizesfor each of the six spatial policy memoranda (first column)the explicitly mentioned input sources. We distinguish between policy-advise reports (second column), advocacy coalition documents (third column) and empirical sources (fourth column). Whereas advocacy coalitions are groups of stakeholders invited to represent their interests, we consider empirical research here to be impartial.
There research models elaborated by one of its departments. There are several state institutes among the sources, including the Centraal Bureau voor de Statistiek (CBS; National Statistics Bureau) and the Planbureau voor de Leefomgeving (PBL; Netherlands Environmental Assessment Agency). Empirical references in the policy documents include articles and international organization reports (e.g. OECD, FAO, World Bank) but almost no independent academic works. None of the 23 papers we found through the Scopus database was referenced in the policy documents.
We occasionally uncovered clear evidence of chains of influence, such as the Fourth Memorandum, which references societal input of advocacy coalitions organized by the Raad van Advies voor de Ruimtelijke Ordening (RARO; Advisory Council for Spatial Planning) as well as empirical modelling and strategic advice from the RPD. We can distil various empirical sources from the reports of these two organizations. However, most memoranda only implicitly reference empirical data, sources and policy theories. Policy documents frequently refer to other policy documents, such as European Council decisions. Several relevant advisory reports did notyet, at leastexplicitly influence spatial planning memoranda (see small arrows in Figure 4). Figure 4 confirms a Dutch tradition of policymaking with the participation of various advocacy coalitions, known as 'poldering'. 40 Our analysis shows that economic interest groups are more widely represented than environmental groups, except for View on the Beauty of the Netherlands, 41 which specifically focuses on the landscape effects of business sites adjacent to motorways. Typical in the Dutch logistics sector are interest groups like Transport and Logistics Netherlands (TLN), EVO-FENEDEX and Holland International Distribution Council (NIDC). The NIDC was founded in 1987 to promote the Gateway to Europe narrative, promote the Dutch logistics sector abroad and serve its interests in the Netherlands ( Figure 5). Its approximately 300 members include logistics companies and governments. 42 None of these three organizations, however, explicitly lobby on the spatial effects of logistics companies. TLN's lobby targets 7 Dutch ministries, not including the ministry responsible for spatial planning 43 , while EVO-FENEDEX does not mention it either. 44 The data show frequent influence on spatial planning with regard to the logistics complex by the Ministry of Economic Affairs and the Ministry of Infrastructure. Actors like the Rotterdam Port Authority, a public company owned by the Rotterdam municipality (71%) and the national government (29%), also wield significant influence over the policy debate.
Regarding influence by research or policy advice, it is remarkable that the two main critical policy advisory reports about the role of the Netherlands as a Gateway to Europe have been completely ignored in the policy memoranda. 45 These reports conclude, based on empirical evidence, that the heavy transportation function is the least profitable and most polluting element of trade. Therefore it would be more economically advantageous to focus on digitalization and trade-management activities, which are highly profitable, while channelling goods traffic partly through other territories. The reports argue that heavy infrastructure has deleterious environmental effects, decreasing the competitiveness of the Dutch economy. Neither of these reports were received warmly. 46 In an official reaction to the 2016 Rli report, 47 the Minister of Infrastructure asserted, without any evidence, that growing transport volumes are necessary to remain a successful trading country and that state programs are effectively dealing with the issue of added value. This reaction ignored the negative effects of freight transport altogether. Spatial policy memoranda also routinely ignored reports discussing the difficult trade-off between risks and benefits of the mainport policy. 48

Key policy theories in use
From the policy memoranda, we distilled three dominant policy theories on the Dutch logistics complex and its spatial effects. Here, we introduce eachfrom broad to specificwith a brief summary of the critical causality chain: problem → policy goal → policy instrument.
Gateway to Europe or perish The Netherlands must remain a leading trading nation → the country needs to strive for a key position in emerging global value and supply chains = become the Gateway to Europe → public investments in mainports and the hinterland logistics complex are critical.
This argument was particularly prominent in the policy memoranda of the 1980-1991 period. 49 The hinterland logistics complex was conceptualized in the policy documents as a logistics delta and a port-industrial complex. Government investments included a major extension of the Rotterdam Port beyond the existing coastline, named Maasvlakte 2, which was heavily contested by environmental groups. 50 The Betuwe line, a dedicated freight railway costing €4.7 billionfour times the initial estimatewas also heavily contested. 51 Meanwhile, policies actively stimulated private initiatives in hinterland distribution clusters. 52 In recent policies, this notion of the Netherlands as a freight gateway still holds strong. 53 The 1988 Fourth Memorandum of Spatial Planning 54 references evidence from the main infrastructure advisory report, 55 which demonstrates two trends in logistics: growing freight volumes and the rising importance of logistics supply-chain management. Spatial policy was clearly adapted to accommodate the former. While a potential threat to the Dutch trade position was mentioned, no evidence of this was provided in the documents. Nevertheless, in 1986, the Dutch Minister of Infrastructure warned that the Netherlands must not become the 'Jutland of Europe', 56 meaning a peripheral country: a typical example of fact-free framing, since Jutland in Denmark should in no way be considered to be a 'backward' region. Similarly, the 2004 Spatial Strategy repeated the self-declared success story of the Dutch economy, confusing the effects of topography and spatial policy: The delta provided the opportunity to develop ports and efficient transport systems with significant economic opportunities for trade, distribution and related logistics. Direct connections between the large ports (mainly Amsterdam and Rotterdam) and the hinterland became the backbones for economic development. 57 Ample supply of space for logistics as economic necessity To maintain economic growth and avoid unemployment → sufficient land for logistics developments must be supplied → regional and local governments need to use their spatial planning competences to make this happen. This policy theory became popular amid the 2000s decentralization wave. 58 Initially, the supply of motorway locations for logistics was regarded as both a national interest and a concrete policy task. 49 VROM, "Vierde Nota Ruimtelijke Ordening," 41. 50 Wolsink, "Reshaping the Dutch Planning System"; Van Gils and Klijn, "Complexity in Decision Making." 51 Priemus, "Development and Design of Large Infrastructure Projects." 52 I&M, "Structuurvisie Infrastructuur En Ruimte," 83; VROM, "Vierde Nota Ruimtelijke Ordening,[26][27]48, BZK, "Nationale Omgevingsvisie-NOVI," 32. 54 VROM, "Vierde Nota Ruimtelijke Ordening." 55 RPD, "Ruimtelijke Verkenningen Hoofdinfrastructuur." 56 Van Duinen, "Mainport and Corridor." 57 VROM, "Nota Ruimte-Ruimte Voor Ontwikkeling," 14. 58 Ibid., 9.
While the national interest continues, the task was decentralized. Although this remains the norm at the time of writing, the most recent planning memorandum from 2020 emphasizes collaboration with local governments. 59 The alleged necessity of low-skilled job creation has been a constant aspect of the Gateway to Europe narrative, first of all in national policy during the economic crisis years of the 1980s, and later in local planning policies geared towards logistics developments.
None of the planning memoranda included evidence of the effectiveness of decentralization in spatial policy. The supply of land for logistics development was regarded as an evident success, as growth assessments of the logistics complex in hinterland locations over the last decades showed a steady growth of logistics jobs in logistics regions and even steeper growth in the spatial footprint of logistics real estateover 300% since 1980. 60 However, the lingering boxification debate suggests that the environmental and landscape impacts have yet to be sufficiently handled. The job argument became less prominent once it became apparent that many of the low-skilled jobsand even many of the high-skilled jobsin logistics can only be filled with migrant labour due to Dutch labour shortages. 61 Mitigation of the spatial impacts of logistics Negative effects are inevitable in the growing logistics complex → the Netherlands should strive to minimize these effects without curbing growth → innovation and win-win scenarios should be stimulated.
This desired win-win scenario for logistics and the environment has been a mainstay in policy documents for the last 40 years. The most recent memorandum promotes space for both healthy living and more air travel; for both an attractive landscape and sufficient land supply for logistics. 62 This firm but almost naïve belief in the potential of a win-win scenario seems to be rooted in a permanently optimistic attitude toward technology. 63 Negative effects, such as the congestion of transport infrastructure and the growing footprint of logistics activities, are expected to eventually be solved by logistics innovations. Such innovations include synchromodality, which aims for infrastructure-, warehouse-and vehicle-use optimization through information sharing among actors in freight transport, and the physical internet, an advanced version of synchromodality with high levels of freight standardization, consolidation and automationstill considered utopian by many experts. 64 Environmental concerns surrounding logistics have been prominent since the 1980s. 65 The motorway panorama policy 66 introduced the idea of building-free zones along certain national motorways. The balance between maintaining open space and the stimulation of distribution and production facilities along motorways, however, remained a regional and local responsibility. 67 As of the most recent memorandum, distribution centres are explicitly linked to cluttering and fragmentation of 'outstanding landscapes', which should be addressed by regional environmental agendas. 68 While policy advisors raised concerns early on over the focus on the Netherlands' distribution function with no consideration of its effects on the Dutch landscape and environment, 69 infrastructure development models of the 1980s showed considerable negative ecological and landscape impacts. 70 Nevertheless, the eventual observation of boxification and motorway landscape disruption was met with shock. 71 Evidence of a successful mitigation of environmental impacts by technology remains scarce; technology's role as a driver of logistics growth, however, has become readily apparent. Teleshopping (the precursor of e-commerce), for instance, was welcomed with interest in the early 1980s and recognized as a positive game-changer starting in the mid-2000s 72and most of the recent growth of logistics land use effectively stems from this innovation. 73 Evidently, logistics is no different than coal in Jevons's paradox: the more efficient its application becomes, the greater its consumption. 74

Discussion
Biased policy narrative 'Gateway to Europe' clearly fulfils Hall's criteria (1993) of a policy narrative. First, the sequence of spatial policies has lent legitimacy to the prioritization of logistics infrastructure development over other public interests, such as the quality of the living environment and landscape. This asymmetrical trade-off was explicitly criticized in a 1980s policy advice: ' … in the followed approach, the production and distribution structure becomes determinant for spatial quality, while the first should be derived from the second.' 75 More than three decades later, the Strategic Environmental Assessment of (the 2019 draft of) the National Strategy on Spatial Planning and the Environment repeated the dilemma of 'large economic opportunities versus large environmental quality threats.' 76 Economic opportunity has consistently carried the heaviest political weight since the 1980s.
Second, national policies consistently use a selective interpretation of Dutch history as a trading nation. The first spatial planning memorandum in 1960 began by stating that 'The foundation of the development of the Netherlands is its location in the focal point of transport routes between the European continent and the world seas.' 77 Although policy memoranda suggest a relationship between the Dutch staple ports of the Golden Age and the current containerized logistics sector of re-export and e-commerce, this sector is, in fact, rooted in the more recent transit function of the port of Rotterdam, made possible by the steamship and the telegraph of the nineteenth century. 78 Such a misrepresentation of history, in our view, is comparable to the Belt and Road imaginary mentioned in Section 2. 79 Third, the Gateway to Europe narrative has created a policy context that is biased to logistics developments despite the availability of alternative policy pathways. The stimulation of domestic exports could have been less environmentally damaging but equally profitable. 80 Beyond spatial policies regarding infrastructure investments and logistics development, the state also used nonspatial instruments. These include subsidies to strengthen the so-called 'logistics top sector', a favourable Dutch VAT law (tax is due only when goods are re-exported from a warehouse) and labour regulations allowing night shifts in distribution centres, in contrast to for example Belgian regulations. The next two sub-sections discuss lessons from the Gateway to Europe narrative with regard to the development and adaptation of policy narratives.

Weaknesses of closed policy narratives
A forty-year period with a rather unbalanced trade-off between logistics and its spatial effects has produced two main weaknesses in the Gateway to Europe narrative. Internally, it has led to a widespread belief in an unrealistic win-win scenario in which the growth of the logistics complex can coexist with environmental protection. Policy theories pertaining to the success of decentralization of difficult spatial planning decisions and technological silver bullets sustain this belief. Externally, it has strengthened at least three strong counter-narratives: i) the Netherlands as a trade-control centre, managing flows not only in the Netherlands but beyond; 81 ii) the circular economy, relying on shorter (regional) and more closed value chains; 82 and iii) the knowledge economy, maintaining an attractive landscape with limited boxification to retain and attract talent. 83 Academic and policy discussions have begun to explore the effects of the COVID-19 pandemic on 'slowbalization', the regionalization of supply chains and the perceived need to transform the mainport policy to meet circular economy and landscape goals. 84 Furthermore, policymakers and logistics sector representatives have recently begun to express their hope that a more circular economy with new logistics concepts will soon change the Gateway to Europe approach. 85 Overall, the Gateway to Europe narrative resemblesrather stronglywhat Throgmorton 86 calls 'abnormal discourse', in which logistics and landscape advocates passionately attack each other instead of constructing common discourse. This tells us that an important element of Throgmorton's storytelling approach is insufficient in the Gateway to Europe narrative: an awareness of differing and opposing views. The hegemonic nature of the mainport and Gateway to Europe policies 87 seems to have prevented them from improving their narratives by learning from conflicting views. A dynamic environment in which storylines can coexist and interactwhich Hajer 88 calls a 'discourse coalition', has been severely lacking.

Limited spatial policy learning
The Gateway to Europe narrative has undergone a policy-learning process 89though it has not been as productive as was possible and necessary. The process has been influenced by exogenous economic shocks (e.g. crises and the integration of the European internal market), but also technological paradigm shifts (e.g. the logistics revolution and the rise of e-commerce platforms). In the words of the head of the national spatial planning agency in 1986: 'Spatial planning must react on big changes: global economy, European economy, unemployment, technological developments in production and distribution, and interaction among people.' 90 The process has also been influenced by advocacy coalitions and, to a limited extent, empirical research. Policymakers must decide on emerging issues with limited evidence on account of their novelty. Nevertheless, more use could have been made of available research in at least two ways. First, had the national government explicitly considered critical research regarding the societal advantages and disadvantages of the logistics complex, 91 it may have adapted its policy narrative and related spatial interventions to mitigate logistics' negative effects. Second, had the national government commissioned more researchincluding forecasts and monitoringinto the spatial effects of logistics when the issue was first raised in the 1980s, it would have had a more substantial base of knowledge on which to make decisions for decades to come. While advocacy coalitions of both the logistics and landscapeenvironmental perspectives were heard over the years, the latter group has been notably less significant, less connected to core policy circles and, in turn, less influential. Logistics interest groups have benefited from infrastructure investments, tax cuts and subsidy programs. Landscape and environmental interest groups only gained occasional compensation projects for ecological damage and a program for motorway panoramaswhich was soon dismantled.
Throughout the policy-learning process, leading spatial-logistics concepts increased in scale, from mainport to Logistieke Topsector Regio (logistics top-sector region), transnational transport corridors and a Logistieke Delta (logistics delta)all obvious examples of framing with language. 92 At the same time, the actual spatial planning of distribution centres scaled down, since it became more and more a responsibility of local governments. This scale diversion is widely regarded as a pressing planning problem: well-informed capital-intensive conglomerates make land deals with rural municipalities desperately seeking funds, unhampered by effective policy guidance from regional or national governments. Furthermore, the missed opportunities to consider empirical evidence, critical views and more balanced advocacy coalitions have turned the Gateway to Europe narrative into a rigid spatial planning storyone that has not shifted its main focus from increased trade volume even in light of what today constitute widely accepted policy goals, such as circularity and the avoidance of boxification. Such strong path dependency is likely to cause a spatial-economic lock-in, 93 in which the rising spatial impacts of logistics are, over time, combined with its declining added value and societal benefits.
In line with Hall,94 we concluded that over the last 40 years, the Gateway to Europe policy narrative has prioritized trade over other societal interests and selectively interpreted Dutch national history to facilitate its desired logistics developments. Contrary to the 'planning as storytelling' approach, 95 the narrative has been unable to address the spatial effects of logistics and learn sufficiently from counter-narratives. Optimistic win-win scenarios, policy decentralization and technological silver bullets prevented policymakers from implementing restrictive policies, instead decentralizing tough spatial choices to local governments, which may find it more difficult to resist land-taking attempts by powerful companies. Gateway to Europe has entailed some spatial policy learning; to a limited extent, economic shocks, technological milestones, academic research and advocacy coalitions have influenced the evolution of spatial policy concepts and instruments. 96 Beyond the disproportional prominence of logistics advocacy groups over environmental and landscape advocacy groups, the use of empirical research has been suboptimal in this policy-learning process. Critical reports pertaining to the Gateway to Europe narrative were structurally ignored by policy memoranda, while research into policy alternatives was never even commissioned. The construction of an open narrativeone that includes accurate spatial effects and is based on research and open discourse coalitionsmay provide a way out of the present spatial-economic lock-in.
It would be highly interesting to see comparative research into the formation of Gateway to Europe policy narratives in other countries, on various governmental levels. To achieve a detailed understanding of policy-learning processes, we suggest that future researchers employ stakeholder interviews and the detailed mapping of lobby networks. on the conceptualization of space and territory in Dutch regional and national planning, with an emphasis on visualization and governance.
Paul Gerretsen is a Director of the Deltametropolis Association, Rotterdam. As head of the non-profit Deltametropolis Association, he leads a network of stakeholders and professionals in design, planning and policy making. He stimulates the debate on regional planning and spatial research, scientific as well as design-oriented.