How Digital Disruptors Influence Multi-Level Governance on the EU: The Case of Airbnb and Amsterdam

(1)

Faculty of Governance & Global Affairs

How Digital Disruptors Influence Multi-Level

Governance on the EU:

The Case of Airbnb and Amsterdam

A thesis designated for the Masters of Public Administration at Leiden University by

Isabella Bremmers-Carrasco 2020

Student Number: s1534823 Supervisor: Prof.dr.ing. A.J. Klievink

(2)

Abstract

Question

To what extent has the sharing economy affected the structural, relational and policy factors that facilitate multilevel governance between Amsterdam and the European Union?

Background

Over the past decade, the explosive growth of the sharing economy has ignited political battles. Airbnb has acquired a marked presence and has influenced the global economy in profound ways. However, there have been mixed responses to its operations particularly in the way that it has impacted on the supply of affordable housing in local communities. Local governments efforts to tame its operations have produced varied results whilst institutions such as the EU gave warmly welcomed these digital innovations; The City of Amsterdam is an example of one European city that has felt the brunt of Airbnb and the city identified to be the case study for this thesis. There is growing discord between the different parties so it is important to grasp where the difficulties lay. This research sheds light on the relationship that exists between the EU and the City of Amsterdam by placing it within the framework of multi-level governance.

Aim

This study offers three hypothesis that seek to identify and examine the influence the structural, relational and policy factors have had on the sharing economy (Airbnb). The aim is to investigate these processes to gain an insight into the interplay between the sharing economy and multi-level governance.

Methods

To review and critically appraise the literature from 2016-2020 to establish if the structural, relational and policy factors that exist between the European Union and Amsterdam have facilitated multilevel governance between the two actors. Literature was collated by using keyword searches in relevant

(3)

electronic bases from 2016-2020. Furthermore, there was a thorough examination of the references in each publicised study. In total 34 pieces of literature were identified that met the criteria. The data from this literature was objectively assessed during the content analysis to uncover information that would confirm or reject the proposed hypotheses.

Results/findings

This research highlighted that structural, relational and policy factors are misaligned. The structural processes displayed in this multi-level governance framework presented various aspects of receptivity to policy initiatives but were generally shadowed by the inflexible architecture that exists within the EU. The relational process exhibited the innovative methods utilised by the City of Amsterdam to assemble bottom-up as a way of permeating their demands and ideas upwards towards the EU level. Policy processes demonstrated were shown to be a complex and a multi-level policy problem. The findings supported the overall predication, but only one of the hypotheses was confirmed.

Conclusion

This research found that there was a governance mismatch between structural, relational and policy process. These factors have been ineffective in generating a true multi-level governance outcome. The process fell short of being mutually complementary and encompassing. This review highlights that further research is necessary to establish how these processes can be harmonised to produce a successful multi-level governance outcome.

(4)

Acknowledgments

I would like to thank my wonderful and loving parents for supporting me through this process. Your encouragement and positive mindsets have been priceless. Without your endless love and guidance none of this would have been possible. This one is for us.

To my close friends I express my deepest gratitude in encouraging me to keep going when I lost motivation and providing a shoulder to lean on when it times got tough.

Lastly, thank you to my supervisor Prof.dr.ing. A.J. Klievink for his input during the course of this thesis.

(5)

The current development and rapid advancement of online sharing platforms have sparked considerable debates concerning its proposed economic and social benefits. The recent establishment of such online platforms is commonly referred to as the sharing economy (hereinafter SE) (Dudas, 2018). The SE is widely described as the “peer-to-peer exchanges for renting goods or services utilising Internet platforms” (Ganapti & Reddick, 2018, p. 77). Platforms that operate in the SE concentrate on the “peer-to-peer economic transactions by facilitating the sharing or renting of space, assets, and labour in real-time (Ganapti & Reddick, 2018, p. 77). The SE has expanded rapidly globally, with its predicted growth to be worth 335 billion dollars in 2025 (PricewaterhouseCoopers, 2015 as cited in Ganapti & Reddick, 2018).

Airbnb is one of the fastest-growing online platforms (Gyodi, 2019), offering up to 4 million rooms globally (Airbnb 2017, as cited in Gyodi, 2019). The increased professionalisation and technological sophistication of the platform have contributed to its profound growth (Parente et al., 2018), and is having a significant effect on urban cities. Not only has it become a fierce competitor and disruptor in the traditional hospitality industry, but it has also triggered a crisis in housing availability and affordability (Nieuwland & Melik, 2018). While Airbnb advertises “itself as providing unique and authentic accommodation and travel experiences” (Nieuwland & Melik, 2018, p. 812), its associated conflicts and challenges have continued to gain traction as its negative impacts have become unsustainable for many cities in Europe (Gyodi, 2019). Subsequently, the debate surrounding these digital platforms has intensified, prompting many “city governments to pass new regulations attempting to control both their proliferation and platform activities” (Aguilera & Artioli & Colomb, 2019, p. 4).

Amsterdam is one European city active in attempting to regulate Airbnb (van der Zee, 2016), and to reduce “its impact on residents and local communities” (Lomas, 2020). Attracting

(10)

over 5.7 million tourists in 2014, Amsterdam was the 7th most visited destination in Europe and 27th globally (European Commission, Directorate General Internal Market, Industry, Entrepreneurship and SME, 2018). The city, which declared itself the first “sharing city” in 2016 (Agyeman & Mclaren, 2016; Malhotra & Van Alstyne, 2014), has placed itself as the hub of innovation for the SE (Oskam, 2019; Augenstein & Bachmann, 2018). Nevertheless, in recent years, the relationship between the City of Amsterdam and Airbnb is shaping up to be an acrimonious one. The concerns are grounded in arguments that Airbnb is “removing housing stock and driving up rental prices, hollowing out local communities and creating other types of anti-social disruptions” (Lomas, 2020; van der Zee, 2016). These issues are spilling over into fierce debates in the European Union (hereinafter EU), establishing “a tug-of-war between cities, platforms” and the EU (Vidal, 2019, p. 1). In December 2019 this came to a head when the French hotel industry lobby bought a case to the European Court of Justice (hereinafter ECJ) arguing that “Airbnb operates as a housing letting agency (short and long term), therefore

French law should prevail to the operations of its business.” (“France 24”, 2019). The ECJ

asserted that Airbnb should not be classified as a real estate agency but rather as an “information society service” (Mehring, 2019). The reading behind this is that Airbnb does not provide an accommodation service but facilitates a platform where customers can select a place to rent where the price is not predetermined by Airbnb (Mehring, 2019). The significance of this ruling has meant it is tougher for European municipalities to administer “tighter restrictions as such services remain regulated under existing EU E-commerce rules” (Lomas, 2020) and it allows Airbnb to operate as normal (Sonnemaker & Ghosh, 2019). This creates further challenges for municipalities across the capital cities of Europe, for example, Amsterdam, who is endeavouring to deal with the fall out of the digital transformation on its neighbourhoods, as Airbnb engulfs popular cities throughout Europe.

(11)

In the context of the EU, this has been a challenge to governance at all levels; there are increased demands from all parties to clarify the rules that govern the SE in Europe (Beltrà, 2016). For cities, the high volume of tourists and “the reallocation of housing units to more profitable short-term stays have caused considerable displacements of local tenants and reduction of the long-term residential stock” (Videl, 2019, p. 1). As a result, many municipal governments are seeking to restrict the number of short-term rentals in order to reduce the problems occurring at the local level (Videl, 2019, p. 1). However, such ‘local’ problems in cities are continuing to be overlooked by the EU. The SE is looked upon favourably by many EU staff and politicians as it boosts economic growth within a competitive single market (Videl, 2019, p. 1). Moreover, the EU is also scrutinising local government regulation on short-term rentals (hereinafter STR) making it more difficult for local authorities to rein in the global tech giants, namely Airbnb (Videl, 2019). Consequently, these dynamics are creating institutional and legal friction between some European cities and the EU in their approach to regulating online STR platforms. The prevalent Multi-Level Governance (hereinafter MLG) framework in the EU is being challenged by the diverse attitudes to the role of the SE in the community. Even though local governments are in closer proximity to the urban inhabitant than EU bodies, the authority of the European Commission and the ECJ means there has been discord on the interpretation of how such “outdated” EU regulations, that were written in times prior to the development of the SE, should operate in contemporary Europe (Videl, 2019, p. 2). Subsequently, the aim of this paper is to look at the role of the SE, (Airbnb) and the Municipality of Amsterdam’s concerns at its proliferation and its impact on MLG in the EU.

(12)

1.1a - Problem statement

The SE is a challenge that has materialised in a multi-level context. Policies that have been developed at one level are creating unforeseen implications for actors working at other levels. As a result, questions of authority and competencies stand unresolved. Systematically, this leads to the questions of how MLG of the SE truly operates and who the participating actors are and what their roles are. While prescriptively, this thesis examines whether the management of the SE can be deemed as an effective illustration of MLG. The social consequences of the SE in urban regions, have led to substantial challenges for local governments at the urban level; they are often left unprepared in both policy and administrative functions. Subsequently, the SE has been selected as the subject matter of this thesis as it encompasses the local, national and European levels.

In light of the SE being a relatively new phenomenon, there is limited knowledge regarding the MLG of the SE and its associated platforms in Europe. A wide range of research papers have analysed these platforms based on their commercial activities, financial models and digital operations. In the paradigm of Airbnb, its effects on the hospitality and tourism industry have been studied greatly but they do not offer an answer to the unresolved friction between the EU and European Municipalities; in particular Amsterdam. Moreover, the broader impact of the SE and its relationship to the European framework has received little attention by scholars from an MLG position. Studies have primarily focused more on its legal implications in its wider context (competition policy) and have not addressed the impact on European cities, the economy and society as a whole.

Additionally, most studies on MLG tend to concentrate on EU integration and the EU-Member State relationships (Bache, 2008; Hooghe & Marks, 2001). This study will primarily look at the role of Amsterdam, as a European municipality and its direct relationship with the EU and the Netherlands as a Member State. It will be investigated through the framework of

(13)

MLG and by examining the issues, which are manifesting themselves at the local level. To accomplish this, it will build on existing research by Dion Curry (2018) and will make use of his new framework for MLG which analyses the structural, relational and policy processes that facilitate MLG (Curry, 2018a). By adopting this approach MLG is no longer presumed to be the ‘independent variable’ as one cannot assume it is a given (Ostaijen & Scholten, 2018). Instead, its pertinence should be examined as one conceptual configuration as it helps provide a thorough apprehension of what is required and why MLG develops in particular situations (Ostaijen & Scholten, 2018). By studying MLG as the dependant variable, it allows for a “more precise theoretical understanding and conceptualisation” (Ostaijen & Scholten, 2018). Consequently, this thesis will seek to reduce the gap in the academic literature and will attempt to locate the responses to the SE in a field that exhibits both “issue complexity and institutional complexity” (Stephenson, 2013, p. 187 as cited in Curry, 2018a). It will investigate how Amsterdam as a European Municipality, functions within this framework where the “governance arrangements are often correspondingly complex” (Curry, 2018a, p. 141). In addition, it aims to contribute to the current debate on how and why local governments influence governance in a multi-level setting.

This is significant as it is important to make clear where the real-life difficulties lay and to expand our understanding of MLG as a functional, systematic and prescriptive concept. Furthermore, comprehension of the aforementioned holds societal relevance as it improves ones knowledge of MLG and therefore helps acknowledge the entanglement of a complex issue that cuts across political structures and policy boundaries. By applying existing theory to emerging real-life challenges new contributions to studies of MLG will be provided. While this improves existing theoretical assumptions, it also adds to the discussion surrounding the SE.

(14)

1.1b - Research Question and Sub-Research Questions

The research questions and sub-research questions guiding this thesis follow as:

Research Question:

To what extent has the sharing economy affected the structural, relational and policy factors that facilitate multilevel governance between Amsterdam and the European Union?

Sub-research Questions:

❖ How does the sharing economy affect the structural factors in Multi-Level Governance?

❖ How does the sharing economy affect the relational factors in Multi-Level Governance?

❖ How does the sharing economy affect the policy factors in Multi-Level Governance?

1.1c – Thesis overview

Following the introduction, the second chapter includes a review of concepts including MLG and the SE. It provides a literature review of MLG and introduces the theoretical framework underpinning this thesis. A critical discussion follows that examines the SE and its context for this study. The work of van Doorn (2019) is introduced to help the reader understand the operations of Airbnb and why it has become a powerful actor in European governance. Chapter three provides a detailed explanation of the methodology used to investigate the research question. Chapter four provides a description of Airbnb in Amsterdam, and a portrayal of the governance structures present in the Netherlands and the EU. Chapter five analyses the collated documents (34) to uncover the structural, relational and policy process that facilitate

(15)

MLG. The final chapter (six), provides a discussion and synthesis of the results, to locate the sharing economy in the European context of MLG and explores the new position of Airbnb. This is followed by the conclusion, which draws upon the main findings uncovered in the research. Lastly, a reflection is undertaken on the strengths and limitations of the study, with suggestions for further academic research.

(16)

Chapter 2 - Multi-Level Governance (MLG)

This section will explore the existing academic literature surrounding MLG followed by Dion Curry’s (2018) framework, where the structural, regional and policy processes are described in further detail. The local level position in MLG will be examined, further adding to the aforementioned framework. This is undertaken as the SE is primarily an urban phenomenon (Vith et al., 2019) Therefore, it is important to ascertain the overarching EU framework as the SE continues to develop in this often complex and changing relationship that prevails between the urban level and the EU.

2.1 - The EU and Multi-Level Governance: Europeanisation

Many MLG scholars discuss the MLG framework in the context of European integration. Inter-governmentalism and neo-functionalist approaches to MLG primarily concentrate on examining it within the mechanism of European integration, the EU’s governmental system and European institutions and their outputs. Other scholars discuss the Europeanisation of the EU. This is relevant because this theorisation creates greater opportunities for local policy actors within the MLG framework (Featherstone & Radaelli, 2003). The latter approach observes the EU to be a powerful and influential system made up of various layers of government and a range of actors (state and non-state), which is the result of competences being divided between the individual layers of government (Featherstone & Radaelli, 2003; Curry, 2018b). As a result, this approach allows for a more detailed understanding of structural, relational and policy processes as power is dispersed upwards towards the supranational organisation (EU) and downwards towards regional and local governments (Curry, 2018b).

(17)

Moreover, Radaelli and Featherstone (2003) identify two different types of mechanisms that are present in Europeanisation; vertical and horizontal. At the EU level vertical mechanisms are notably set apart where policy is shaped, and the domestic level where policy is taken on-board (Featherstone & Radaelli, 2003). On the other hand, at the horizontal level mechanisms examine the involvement of different forms of adjustment to the EU level that are based “on

the market or patterns of socialisation” (Featherstone & Radaelli, 2003 p. 41). Additionally, in

horizontal Europeanisation, there is no pressure to adhere to EU policy models, where a high level of policy “descends into the domestic policy arena as in a hierarchical chain” (Featherstone & Radaelli, 2003 p. 41). Instead, outcomes are triggered by “choice and the diffusion of ideas and discourses about the notion of good policy and best practice” (Featherstone & Radaelli, 2003 p. 41). Bache and Andreou (2013) stress that Europeanisation is an approach which observes, “what is ‘coming down’ from the EU” (Bache & Andreou, 2013, p. 3), in terms of the essence and the force of a particular instrument (Bache & Andreou, 2013, p. 3). Europeanisation also considers how this ‘fits’ into and is conciliated by domestic actors/circumstances (Bache & Andreou, 2013).

Nevertheless, it should be understood that the EU is a convoluted supranational body where policy-making is scattered across a wide range of actors (Grisel & van de Waart, 2011). These actors include “politicians, officers, policy networks, interest groups, and in some cases the wider public” (Grisel & van de Waart, 2011, p. 11) who work together at the local, regional, national and European levels to coordinate and determine policy within a sector (Grisel & van de Waart, 2011). Usually, the collaboration between the aforementioned groups to manage a prevalent issue is called governance (Grisel & van de Waart, 2011; Armondi, 2020). As actors are sited at various levels within the EU (Grisel & van de Waart, 2011), the concept of MLG has arisen as a means to understand the positions of authority and power between the domestic

(18)

and supranational levels as well as to comprehend how they work together (Hooghe & Marks, 2001; Grisel & van de Waart, 2011).

Stephenson (2013) stipulates, “the use of MLG in an endeavour to offer basic ideas to a scattergun approach to the formulation of policies, where numerous actors (individuals and institutions) partake at political levels, from the top (supranational) down (local)” (Stephenson, 2013, p. 817). Even though there is geographical separation, MLG signifies that there are interconnected linkages that are interwoven at these different points (vertical or horizontal) (Stephenson, 2013; Featherstone & Radaelli, 2003) with engagement and influence at each level but none superior to the other: the result of a “mutual dependency through the intertwining of policy-making activities” (Stephenson, 2013, p. 817). Moreover, Ostaijen and Scholten (2018) also emphasised the evenly balanced relationship of various levels in participation and decision making in governance situations. (Ostaijen & Scholten, 2018, p. 6). As a result, in a true MLG setting “policies at various levels are synchronised without a clear hierarchy” (Ostaijen & Scholten, 2018, p. 7). Armondi (2020) further postulates that there is mutual loyalty and interest that prevails between the distinct levels of government and various institutions when trying to reach common goals (Armondi, 2020).

However, this is often not the case as MLG is repeatedly at odds “with the reality of blurring of jurisdictions, dysfunctional coordination, overlapping functions, excess of regulations, or no traditions of cooperating” (CORA, 2013, p. 87 as cited in Armondi, 2020). Grisel and van de Waart (2011), also assert that there are an increasing number of fields where jurisdictions overlap resulting in a whole scope of actors across levels, working together to complete a task or deliver a specific service. Kern (2007) specifies that between these different actors, an emergence of partnerships, networks, alliances and functional associations have been formed. These arrangements are flexible and will (hopefully) be constructed in a manner to

(19)

attain the best performance (Kern, 2007). However, these actors do not always act in a “coordinated and collaborative way” (Armindi, 2020, p. 25).

Hooghe and Marks (2003) discuss the aforementioned by defining two forms of institutional arrangements: Type I and Type II MLG. Type I is similar to federal and intergovernmental arrangements (Chowdhury & Wessel, 2012). These arrangements are identified by general-purpose jurisdictions that are somewhat rigid (with reference to institutional architecture) and tend to be bounded (Hooghe & Marks, 2003; Curry, 2018b; Chowdhury & Wessel, 2012; Stephenson, 2013). Membership is static at the limited number of levels and is non-sectional (Hooghe & Marks, 2003; Curry, 2018b; Chowdhury & Wessel, 2012; Stephenson, 2013). Finally, Type I is identified to have only one related jurisdiction/authority at each level (Stephenson, 2013; Hooghe & Marks, 2003). It is modelled on “the simplistic nature of state control and the exertion of authority in a unitary state” (Stephenson, 2013, p. 821). In contrast, Type II is characterised by task-specific jurisdictions operating at different levels in a flexible manner (Chowdhury & Wessel, 2012). These jurisdictions are described as fluid. Members change depending on issues or policy across unlimited levels (Stephenson, 2013; Hooghe & Marks, 2003). Jurisdictions “often align along policy lines, overlap at numerous different levels, and are not limited to simple territorial distinctions” (Curry, 2018b). Type II can be thought of as a “layered system of co-existing levels of authority – a complex pattern of transnational, public and private institutional relations with overlapping competencies” (Stephenson, 2013, p. 821).

2.2 - Multi-Level Governance: A New Framework

Curry (2018b) states that while Hooghe and Marks’s (2001) typology is useful, it assumes that factors fall into a one-dimensional spectrum, which infers simplicity. However,

(20)

this does not necessarily happen in practice. Curry’s (2018a) approach to examining MLG differs slightly as it “focuses on the factors that go into assessing the nature of MLG” (Curry, 2018a, p. 145). As governance happens at various levels, efforts to coordinate solutions may encounter challenges that reveal themselves in structural, relational and policy ways (Curry, 2018a). As a result, governance structures, policy processes, and actors can be assessed according to these principles as it allows for a “more sophisticated, nuanced framework for types of governance” (Curry, 2018b, p. 145).

The three types of processes (structural, relational and policy) “can be mutually reinforcing, contradictory or separate” (Curry, 2018a, p. 144). If these elements are “mutually reinforcing structures that are supportive of MLG-type processes will develop” (Curry, 2018a, p. 144). This provides actors with more space to navigate policy outcomes in a multi-level system (Curry, 2018a) and be used in a way that upholds “multi-level solutions” (Curry, 2018a, p. 144), where these policies could possibly “produce solutions that make use of structures and relations in a multi-level way” (Curry, 2018a, p. 144). However, when these measures function in a conflicting manner, disjointed and decoupled, governance can occur (Curry, 2018a) where actors, institutions and policies engage in discord (Curry, 2018a). Curry (2018a) also notes that decoupled governance can occur when processes function separately. For decoupled governance to occur there must be “little coordination between actors, institutions and processes” (Scholten 2013 as cited in Curry, 2018a, p. 144).

In addition, Curry (2018a) asserts that his breakdown of MLG when analysing the governance of an issue at different levels relates to Hooghe and Marks’ (2001) initial typology. He also says that governance that leans to the left of Table 2.2c can be representative of Type I MLG, while governance that pivots to the right can be deemed to be closer to Type II MLG (Curry, 2018a). Though, there are various governance outcomes as responses mark different structural, relational or policy factors (Curry, 2018a). These responses depend on whether the

(21)

MLG issue operates at multiple levels and for how long. For instance, there may be a shift of power at the centre to the urban/local level. This does not necessarily “improve the policy’s ‘multi-levelness’ in a governance sense” (Curry, 2018a, p. 145). If EU principles are deficient with solutions created at a different EU-policy level, it will not generate a true MLG response (Curry, 2018a). Curry (2018a) notes that the status quo will endure, granted that there is “no change in structural, relational or policy factors” (Curry, 2018a, p. 145). However, assuming that there is a change in a number of factors, it could possibly produce a governance mismatch. This is where the various levels function “more or less independently” (Curry, 2018a, p. 145). Nonetheless, decoupled governance may also occur, where “different levels may operate at cross purposes” (Curry, 2018a, p. 145). As a result, a real MLG response must incorporate “complementary, mutually supporting shifts in all three factors” (structural, relational and policy processes) (Curry, 2018a, p.145). Therefore, Curry (2018a) puts forward three possible options for the EU and Member States’ policy responses to an MLG issue, similar to the Intra-European movement. In the first scenario, the structural, relational and policy factors are clearly and obviously separated across and within levels, illustrating that MLG is not present (Curry, 2018a). The second scenario concerns the structural, relational and policy factors being misaligned across and within levels, which reveals that there is governance mismatch (Curry, 2018a). Thirdly, interdependent, mutually reinforcing and complementary structural, relational and policy factors across and within levels culminate in MLG (Curry, 2018a). These conflicting governance responses on the three spectrums might display disjointed governance, “while independent, non-reinforcing responses indicate decoupled governance” (Curry, 2018a, p. 146). These processes can be further broken down into 6 sub-categories (Curry, 2018a). Structural factors may be complex resulting in fragmented policy-making (Curry, 2018a). In addition, relational factors like hierarchy might also make coordination arduous as different actors seek out power (Curry, 2018a). Finally, various actors often have “competing, conflicting or shifting

(22)

policy interests and goals” (Tasȿan-Kok & Vranken 2011, p. 16 as cited in Curry, 2018a, p. 144). Curry (2018a) maintains that the aforenoted process allows for a more “granular way of analysing the factors influencing MLG” (Curry, 2018a, p. 144). This mismatch may occur within the “structural, relational or policy factors are at a higher level, where there is a

mismatch between the processes and their intended outcomes” (Curry, 2018b).

2.2a - Structural/Institutional Factors

Curry (2018a) uses Table 2.2c (p. 144) to explain how these factors affect the process of MLG and how it can be influenced by structural/institutional factors. He breaks these factors down into two categories: rigid or flexible. The degree, to which the institutional structures are rigid or flexible, depends to a great extent “on the strength of the legislation underpinning the institutional bounds of actors’ jurisdictional powers” (Curry, 2018b, p. 111). Rigid institutional structures can be characterised as being “binding to the actors involved” (Curry, 2018a, p. 144). This is often, “but not always, tied to the threat (or not) of sanctions for inaction (or negative action)” (Curry, 2018a, p. 144). For example, once a direct jurisdiction over an issue has been established between Member States, these measures will remain relatively stable over time (Curry, 2018a). On the contrary, flexible institutional structures are “non-binding, a ‘soft law’ forms that is more responsive to contextual and policy-specific factors” (Curry, 2018a, p. 144). Zelano (2018) affirms that “soft law is based on the mutual trust between member states” (Zelano, 2018, p. 101). Examples of soft law mechanisms used in the EU are: The Social Dialogue (SD) and Open Method of Coordination (OMC) (Zelano, 2018). Moreover, structural/institutional factors also concern the jurisdictions that make up the policy area (Curry, 2018a). The jurisdictions may be multi-purpose, “geographical or cultural entities that exist across policy areas” (Curry, 2018a, p. 144; Curry 2018b), or policy-specific; meaning that the jurisdiction is distinct to that particular policy which is becoming an “increasingly common

(23)

configuration in EU settings” (Curry, 2018a, p.144; Curry, 2018b). For instance, an MLG problem that has rigid institutional structures may be associated with issues on the urban/local level that are mostly “non-binding and flexible in nature” (Curry, 2018a, p. 146). Therefore, the everyday policy methods that are used to deal with the consequences on the local/urban level are passed down to Member States, who in turn delegate a certain amount of authority to the local level which creates a “more flexible approach” (Curry, 2018a, p. 147). When these two extensive categories come together, they generate an institutional environment in which governance prevails along a spectrum. On this spectrum “there can be clear lines of authority

(in a rigid structure, general-purpose jurisdictions), or one with indistinct lines of authority

(flexible, policy-specific approaches), and any variation in between” (Curry, 2018a, p. 144).

2.2b - Relational factors

In addition, MLG can also be affected by relational factors. These aspects influence how “actors (governmental, quasi-governmental or non-governmental)” (Curry, 2018a, p. 144) collaborate with each other within and between the different levels (Curry, 2018a). Relational factors can be broken down into two categories: hierarchical and heterarchical relations (Curry, 2018a). Hierarchical relations follow “a clear chain of command” (Curry, 2018a, p. 144), whereas heterarchical relations are characterised by “networks in which actors work more collaboratively” (Curry, 2018a, p. 144). Curry (2018a) asserts that in fields where “higher governmental levels have binding power, processes will be more likely to operate in a hierarchical manner” (Curry, 2018a, p. 144). In contrast, heterarchical practices exist at the urban/local levels (Curry, 2018a). Moreover, hierarchical relations can be distinguished by their top-down relationships, which means that policy decisions are normally transferred down by “higher levels of government to be implemented by lower levels of government” (Curry, 2018a, p. 144). Heterarchical relations, however allow for a transfer of policy ideas “where

(24)

processes may originate at lower levels and permeate upwards” (Curry, 2018a, p. 144). For example, an EU directive supplies an explicit hierarchical structure in which urban municipalities and Member States have to conduct themselves to the above directives (Curry, 2018a). However, even if there is a clear hierarchy (such as an EU directive), the process controlling “integration, housing and often social policy” (Curry, 2018, p. 151) in the EU is conducted in a more heterarchical manner (Curry, 2018a). This involves several actors with limited or no hierarchical supervision from the higher echelons of government (Curry, 2018a).

Furthermore, relational factors are determined by formalised and informal relations (Curry, 2018a). Formalised relations are standardised and stable, whereas informal relations are more fluid and ad hoc (Curry, 2018a). These types of relationships “operate in terms of context, policy response, actors involved or other factors” (Curry, 2018a, p. 144). He states that actors may concentrate on using fewer formal methods to build agreement on policy approaches rather than focusing on “formal procedures such as court challenges” (Curry, 2018a, p. 151). These informal structures are intended to create action in the lower levels, when issues become too prominent and rise up the political ladder; the effectiveness of these informal structures ability to generate solutions is substantially reduced (Curry, 2018a). As a consequence, even though urban/local authorities may engage in a range of governance systems in various policy areas, “there is a difference between participation or involvement and true influence or governance” (Curry, 2018a, p. 150).

However, in some cases, local involvement may be prominent when the effect of the problem has an impact on the urban level. Many of the urban level issues, for example housing, require decisions to be made by local authorities, without Member State support (Curry, 2018a). This may produce a policy response that is dynamic while at the same time circumventing or “disrupting more stable, formalised relations networks between levels” (Curry, 2018a, p. 150). This is further noted by Ostaijen and Scholten (2018), who state that traditional top-down

(25)

governance may be challenged from the bottom up; if the issues become salient among member states. However, this is contingent on the matter being deliberated, as shifting agendas may at times clash or coincide (Ostaijen & Scholten, 2018). The interaction between the Member States and local government with European institutions at a horizontal level produces various styles of governance function in a setting of fluctuating alliances (Zelano, 2018). As a result, relations appear on a spectrum, for heterarchical connections/policy, the response can “vary from a highly inclusive one or a more fluid and informal one” (Curry, 2018a, p. 144). Formalised relations tend to be an "exclusive one apparent by hierarchy and stable between actors” (Curry, 2018a, p. 144).

2.2c - Policy Factors

Moreover, MLG is also affected by policy factors that can be classified into two groups: discrete and holistic (Curry, 2018a). This is determined by the structural and relational factor which when joined together can create a policy setting (Curry, 2018a). A discrete policy response “may be discrete and compartmentalised, where problems are treated separately with clear lines of demarcation” (Curry, 2018a, p. 144). On the contrary, holistic policy responses deem every “issue as part of a larger whole covering a broader policy area” (Curry, 2018a, p. 144). However, it should be noted that both approaches require coordination but separate reasons (Curry, 2018a). Discrete policy responses call for coordination in order to “clarify and delegate roles and responsibilities within specific issues” (Curry, 2018a, p. 145), while holistic policy responses demand coordination to secure harmonisation across varying issues (Curry, 2018a). Nevertheless, a lack of coordination can arise between different levels, for example, “governmental levels having different policies on an issue” (Curry, 2018a, p. 145). A lack of coordination can also develop between levels, where for instance a “governmental and non-governmental actor has different roles, policy goals and approaches” (Curry, 2018a, p. 145).

(26)

Curry (2018a) states that the level of coordination could be bound “to whether competence over that policy issue is exclusive or shared, and whether those creating the policy are the same as those implementing the policy” (Curry, 2018a, p. 145). Lastly, uncoordinated reactions may arise if a policy issue is disputed, “either through new or unclear jurisdictional bounds or through attempts by actors to take over or relinquish ownership and control of the policy” (Curry, 2018a, p. 145).

In addition, the policy issue can also be described as complex or uniform (Curry, 2018a). Complex issues are crosscutting and “have incomplete, contradictory or changing solutions” (Curry, 2018a, p. 145) or have a wide range of viable options but with no obvious best choice (Curry, 2018a). Furthermore, different levels might not consistently work jointly together, leading policy responses that might misalign and conflict with each other (Curry, 2018a). As a result, a governance mismatch might arise in the policy area (Curry, 2018a). Moreover, the problem itself might become entangled in other governance issues like existing social, demographic and employment problems in a country (Curry, 2018a). A uniform policy issue can be defined as “discrete and straightforward” (Curry, 2018a, p. 145). Consequently, policy approaches can be located on a “spectrum from chaotic (uncoordinated, contested, complex and non-state controlled), to controlled (coordinated, uncontested, straightforward, state-controlled policy)” (Curry, 2018a, p. 145).

(27)

Table 2.2c

Note. Reprinted from “Intra-European Movement: Multi-Level or Mismatched Governance?”, by Curry, D., 2018a, p. 144, Between Mobility and Migration, Cham,

Switzerland: Springer International Publishing AG

2.3 - Multi-Level Governance: The ‘Local Turn’

Further building on 2.2a, a deeper understanding of the local level is needed in order to assess if true MLG is occurring in this case, and if not, what situation is occurring. This is significant as most policy response happens at these lower levels. Therefore, it is important to grasp how this interaction inserts into the MLG framework to enable the assessment of structural, relational and policy processes.

Metropolitan areas are increasingly being impacted by EU legislation in various policy fields (Kharchi, 2017). In most cases, these cities authorise EU programmes, administer EU policies and change local laws when the national level commits to new EU legislation (Kharchi, 2017). This interconnected and extensive relationship is attributed to the expansion of the EU over the past decades (Kern & Bulkeley, 2009). Consequently, urban decision-makers are

(28)

raising questions about the “local implications of the single market and monetary union” as the EU grows increasingly interwoven (Wolffhardt & Hamedinger, 2010 as cited in Kern & Bulkeley, 2009). There is nearly no field of “local or regional policy that is not affected by European legislation” (Kern & Bulkeley, 2009, p. 313).

Kern and Bulkeley (2009) postulate that these “these top-down and bottom-up dynamics” (p. 312) are not restricted to describing the relationship between member states and the EU, but are also applicable to explain the connections between local governments and EU institutions (Kern & Bulkeley, 2009; Zapata-Barrero et al., 2017). Bottom-up governance is described as local governments taking a leading role in agenda-setting and generating policy solutions that involve policy actions at higher levels of governance (Ostaijn & Scholten, 2018). While top-down governance is distinguished by “a hierarchical relationship in which the highest level steers the overall governance response for actors from all other involved layers” (Ostaijn & Scholten, 2018, p. 7). Policy levels below are generally presupposed in terms of policy implementation and are not involved in the formulation of policy (Ostaijn & Scholten, 2018, p 7). The EU has the absolute “responsibility and authority to enforce the measures decided” (Scholten & Ostaijen, 2018, p. 8).

As an example, Kern and Bulkeley (2009) comment on how the EU’s legal and financial instruments influence the local government (top-down). A noticeable example of this is the environmental and sustainable development policy where domestic policy-making in this policy area is upstaged by EU decision-making as it has become more important (Kern & Bulkely, 2009). In addition, local governments also demonstrate bottom-up vertical structures in influencing EU decision-making: indirectly by lobbying at the national level and directly at the European level through networks and national associations.

As a result, scholars are increasingly identifying a “process of urban Europeanization” between the local level of EU Member States (Kern & Bulkeley, 2009, p. 312), which

(29)

encourages new horizontal partnerships, networking and community involvement in policy formulation and decision-making. This cooperation between the EU and local government is significant as they “jointly develop innovative solutions for problems with which they are similarly confronted” (Kern & Bulkeley, 2009, p. 313). Policy arenas are ceasing to be constricted to a certain level as local actors might work closely “with representatives of national ministries and the EU commission” (Kern, 2007, p. 3). Even though these developments may give rise to new opportunities for local authorities to engage in their interests at both the national and European level (Kern, 2007, p. 3) it also makes the coordination of effective MLG more difficult.

2.4- Sharing Economy (SE)

In light of the SE being a relatively new phenomenon, it is important to grasp a deeper understanding of the academic debate surrounding the SE and the externalities, to ascertain its challenges for actors at the local/urban level. Moreover, it is pivotal to understand the workings of the SE; its complexity in an equally complex framework.

2.4a - The Sharing Economy: A Global Phenomenon

The relationship between digital-based platforms in the SE and the European municipalities has become a contentious one (Videl, 2019). The SE surfaced in 2008 in the U.S. out of the “technological possibility and economic necessity” (Schor & Cansoy, 2019, p. 51) that had emerged in the late 2000s. Although the rapid expansion of the SE has had an extensive impact on society, there has yet to be a consensus of how to define what the SE is entirely (Acquier et al., 2017). This lack of agreement is due to the SE embracing “heterogeneous practices and sectors” (Acquier et al., 2017, p. 3), which encompasses a vast scope of organisational formats extending from for-profit to non-profit initiatives (Acquier et

(30)

al., 2017; Parente et al., 2018; Gössling & Hall, 2019). In addition, the label SE, in general, can extensively be used to include sharing-based business models and is applied correspondingly to the terms “access economy, collaborative consumption, and on-demand services” (Parente et al., 2018, p. 54). It is visible in many industries such as the business-to-consumer (B2C) sector, in particular, Zipcar (see Table 2.4a) (Pushchmann & Alt, 2016; Parente, et al., 2018). This thesis will use the Parente et al. (2018) system to characterise the SE and to explain how SE business has proliferated globally (see Table 2.4a).

Despite the lack of agreement on how the SE can be defined, the definition provided by Oxford English Dictionary, describes the SE is “a financial model where goods or services are

shared peer to peer, either free or for a fee, facilitated by a community on-line platform”, is

useful for this analysis (“Lexico.com”, 2020). The aforementioned definition and the description provided by Acquier et al. (2017) both recognise that platforms in the SE can either utilise monetary benefits and/or be non-monetary. As this covers a breadth of platforms, this study will examine a for-profit sharing platform to limit the scope of this investigation. Subsequently, it is important to further improve our definition of monetary platforms in the SE in light of the preceding statements failure to acknowledge how micro-businesses can develop out of the SE as individuals are presented with an opportunity to advertise spare resources which belong to them (Seely & Balogun & Grimwood & Cromarty & Sandford & Foley & Brien, 2018; Acquier et al., 2017). Parente et al. (2018) note this as a characteristic of the SE (see Table 2.4a). As a result, the definition of the SE utilised by the trade association for the SE in the United Kingdom (UK), Sharing Economy UK, is used to supplement the Oxford English Dictionary. The Sharing Economy UK, describes the SE as a phenomenon that “uses internet

technology to create platforms which groups of people and organisations can access, to

facilitate trade or sharing, in order to reduce the need for ownership” (Seely & Balogun &

(31)

Table 2.4a

Note. Reprinted from “The Sharing Economy Globalization Phenomenon: A Research Agenda”, by Parente, C., Geleilate, G., & Rong, K., p. 54, Journal of International

Management, 24(1), p. 54

Subsequently, discussions have been focused on the different perceptions of sharing (Etter et al., 2019). Scholars like Belk (2014) postulate that the sharing which is taking place in the SE is not truly sharing but a kind of pseudo-sharing, “whereby people coordinate the

acquisition and distribution of a resource for a fee or other compensation” (Belk, 2014, p.

1597). This type of sharing occupies the sections between “marketplace exchange and true

sharing” (Belk, 2014, p. 1597; Etter et al., 2019), which creates a “commercial relationship between the owner and the user” (Belk, 2014, p. 1597). A well-known example of this is

Airbnb, which is regularly discussed as being part of the SE (Dreyer et al., 2017, p. 89). The SE is a new kind of market that offers people an opportunity “to rent to or freely share these

assets with others” (Schor & Cansoy, 2019, p. 52). This is formulated on the assumption that

there “is a large amount of excess capacity in the assets held by ordinary households (rooms,

cars and durable goods like tools)” (Schor & Cansoy, 2019, p. 51). This is further reinforced

by Frenken and Schor (2017), who speak of the SE as the notion of sharing “idle capacity” to create “on-demand personal services” (Frenken & Schor, 2017, p. 6) as the key to the current debate concerning home-sharing platforms (Frenken & Schor, 2017). For example, if a homeowner has an unoccupied bedroom (an underutilised asset) and decides to take a vacation or goes on a business trip, that vacant house or room can be treated as “temporary idle capacity” (Frenken & Schor, 2017, p. 6). However, a micro-business has the possibility to develop when

(32)

one may want to purchase a “second home and rent it out to tourists permanently” (Frenken & Schor, 2017, p. 5). This would “represent running a commercial business operation” (Frenken & Schor, 2017, p. 5). Platforms like Airbnb allow homeowners to advertise and disguise these “spare resources” by turning them into a sort of micro-business (Seely & Balogun & Grimwood & Cromarty & Sandford & Foley & Brien, 2018; Acquier et al., 2017). As a result, this phenomenon has blurred the line between business and consumer (Eckhardt, 2019; Vith et al., 2019) and has become a defining factor of SE as ordinary people have become “alternative suppliers of goods and services which have been traditionally provided by long-established industries” (Zervas & Proserpio & Byers, 2017, p. 687). Acquier et al. (2017) also affirm that SE platforms have thwarted the “established markets and bureaucratic organizations that build on asset integration like: transportation, accommodation, and financial services” (Acquier et al., 2017, p. 3).

Moreover, these ‘sharing’ undertakings are not recent; however, they have predominantly occurred between people who are familiar with each other and/or between members who make up social networks, which “were mostly informal” (Schor & Cansoy, 2019, p. 53; Frenken & Schor, 2017; Etter et al., 2019). These informal social networks were not greatly monetised (Schor & Cansoy, 2019). The SE has transformed these relationships into formal large networks that were no longer based on social ties but by transactions made from an unknown stranger to another unknown stranger (Schor & Cansoy, 2019). Consequently the SE has monetised the consumption of excess capacity, arguing that it would bring about financial and social benefits (Schor & Cansoy, 2019; Ritter & Schanz, 2019). Sutherland and Jarrahi (2018) concur with the aforementioned authors by stipulating that the sharing of goods and services is not a new phenomenon; however, the digital platforms and other large-scale mediating technologies that facilitate this large-scale sharing are fairly recent (Sutherland & Jarrahi, 2018; Etter et al., 2019). The establishment and growth of the SE has been enabled by

(33)

the rise of efficient and scalable technologies that have facilitated the activity of acquiring, providing or sharing access to goods online (Sutherland & Jarrahi, 2018; Geissinger, 2020). Frenken (2017) agrees with this study, stating that the social collaborative networks of the SE business are frequently connected to the technologies they operate on. Therefore, it can be acknowledged that the SE is economically motivated. This is in line with the argument of the aforementioned scholars that to access the idle capacity is facilitated through remunerative transactions, which are not socially motivated sharing (Eckhardt, 2019; Etter et al., 2019; Belk 2010; Acquier et al., 2017; Frenken & Schor, 2017).

Furthermore, platforms operating in the SE are highly influenced by network effects (Parente et al., 2018). This occurs “when the value of a product or service increases as more users utilise the platform” (Parente et al., 2018, p. 53). It in turn, expands the network and increases the reach of the platform (Parente et al., 2018). Parente et al (2018) assert that Airbnb creates value on its ability to capture idle capacity as well as being able to “access, store, and provide information regarding lodging solutions to consumers, while also serving as a global marketplace for providers to display their wares” (Parente et al., 2018, p.54). Consequently, SE firms, like Airbnb, are able to spread globally because they make use of “local user's resources to supply local demand” (Parente et al., 2018, p. 54), which cuts costs when expanding internationally. This approach differs from traditional businesses that rely heavily on acquiring assets in other countries (Parente et al., 2018).

Additionally, SE platforms also value simple organisational structures “mainly divided into platform technology, operations, marketing, and customer service” (Parente et al., 2018, p. 53). The most expensive investment for SE firms when they expand internationally is the “platform coding configuration for local adaptation” (Parente et al., 2018, p. 54), resulting in SE firms coming into conflict with “local players and regulators. The SE approach that “expansion strategy prioritizes market share gain and first-mover advantages despite local

(34)

issues related to competitive retaliation and regulatory enforcement” (Parente et al., 2018, p. 55) derives from a platform-based business model which relies on the excessive adoption of its product by users, creating that international “buzz” for the firm (Parente et al., 2018). For instance, Parente et al. (2018) comment on the “shock and awe strategy” (p. 55) undertaken by Uber to create consumer enthusiasm which benefits its lobbying efforts to formalise its enterprise (Abboud & Wagstaff, 2015 as cited in Parente et al., 2018). This strategy has led to cities around the globe becoming both loci of action and central actors (Brandtner & Suárez, 2017) in the debates over the nature and organization of the SE (Kornberger et al., 2018). Therefore, city administrations and governing authorities have found themselves, key players, in the SE with a need for tactical involvement. The above definitions and discussion have highlighted how Airbnb can be characterised as a firm operating in the SE. Furthermore, Parente et al. (2018) note that the consumer pays for the service, which further narrows the scope of this thesis; the role of for-profit sharing platforms will be examined as the profit motive behind the case study (Airbnb) is built on monetised sharing for its users (Acquier et al., 2017). Parente et al. (2018) argue that SE businesses have been able to spread internationally by using “an aggressive market-seeking approach not bounded by cultural or economic discrepancies” (Parente et al., 2018, p .52). As a result, the SE phenomenon has counted on social support and involvement of its global users thereby allowing it to gain significant prominence both internationally and locally (Parente et al., 2018).

2.5 – Airbnb: A Disruptor to European Governance

It has been argued in 2.4 that Airbnb is part of the SE and have been allowed to expand their operations exponentially globally. This has enabled the organisation to become a powerful and influential agent in society. Van Doorn (2019) suggests that Airbnb should not just be considered as just a sharing platform but actually as a “new urban global organisation that is

(35)

changing the landscape between market, state, and civil society actors” (van Doorn, 2019, p.

1815). He also argues that Airbnb has now taken on a role in agenda setting by becoming an urban regulatory entrepreneur, co-shaping “the terms of current and future policy debates

pertaining not just to home sharing/short-term rental but also to the very fabric of city life”

(van Doorn, 2019, p. 1815). Airbnb is a growing institutional power, that seeks to partner with municipal governments in order to take an active role in shaping policy debates and regulations (van Doorn, 2019). Airbnb has increased its role in urban governance seeking to gain “power that would secure them increasing control over particular functional arenas (e.g. tourism, housing, urban planning) which traditionally are governed by state actors” (van Doorn, 2019, p. 1816). Airbnb accomplishes this by mobilising its users under the umbrella of the ‘Airbnb Citizen’ which operate as a collective force (van Doorn, 2019). This initiative forms part of an advocacy channel that has overseen projects such as “sharing Airbnb community data, and releasing a Policy Tool Chest” (van Doorn, 2019, p. 1810). By undertaking the aforementioned enterprises, Airbnb seeks to improve its democratic legitimacy (Smigiel, 2020). However, Smigiel (2020) argues that this has by no means been a one-dimensional process and has been facilitated by the EU Commission (Smigiel, 2020). The legal framework visible in the EU “has supported and strengthened Airbnb’s position” (Smigiel, 2020, p. 255). This predominate economic perspective taken by the EU Commission strongly prevails and it is seemingly more important to the Commission than the concerns held by local governments (Smigiel, 2020). In addition, Smigiel (2020) also notes that Airbnb’s influence has been allowed to expand as it takes advantage of “its Irish license in order to outplay the data restrictions and tax regulations of other EU countries” (Smigiel, 2020, p. 255). Subsequently, the profound growth of Airbnb has allowed it to become a policy-maker in its own right.

(36)

2.6 – Hypotheses

This section will outline the prospective hypotheses that will be confirmed or rejected in Chapter 5. Based on Curry’s (2018a) framework on MLG, this study hypothesises that there has been a governance mismatch as structural, relational and policy factors are misaligned across and within the different levels.

Hypothesis 1: The structural factors that facilitate Multi-Level Governance have lessened the impact of the sharing economy.

Hypothesis 2: The relational factors do not restrict Multi-Level Governance when trying to manage the sharing economy.

Hypothesis 3: The policy process of Multi-Level Governance has negatively affected the impact of the sharing economy

(37)

Chapter 3 - Methodology

This chapter provides an overview of the research design and the methods employed to investigate the research question. To begin with, a discussion on the rationale for selecting a single case study design is undertaken, followed by the outlining of the case section strategy and the defining of scope conditions to streamline this research. Next, an explanation on the case selection strategy is provided to ascertain why and how Amsterdam was selected to be the focus of this investigation. Following this, an in-depth explanation on the data collection method and strategy is given as well as a description of the data analysis technique used to examine the phenomenon of interest. Subsequently, the independent and dependant variables are operationalised (Tables 3.5 and 3.5a) in order to adequately assess the proposed hypothesis. The operationalised variables are displayed next to their definition and indicators. Lastly, an evaluation of the reliability and validity of this thesis is performed.

3.1 - Research Design: Single Case Study

This paper will utilise a single case study design (Airbnb in Amsterdam 2016-2020) by which multiple pieces of evidence are studied to uncover observations within a single unit (Toshkov, 2016). This approach is suitable for this particular research as it aims to study the phenomenon of the SE in the EU as a whole and between the municipality of Amsterdam as a specific case. This will be accomplished by attempting to reveal the outcomes of MLG (MLG, disjointed and decoupled). Specifically, this thesis will examine the outcome of MLG rather than the process. The outcome is defined as MLG, disjointed governance and decoupled governance, which the research aims to uncover by applying an outlined theoretical framework. Although MLG is not a new concept, the application of Curry’s (2018) framework in academic research has been rather limited; therefore, this thesis will help to sharpen new theoretical ideas about MLG (Toshkov, 2016). The “observations” are the structural (referring to the institutional

(38)

factors), relational (collaboration between different actors) and policy factors (the policy setting). These components affect the processes of MLG in the EU. The application of a single case study design is embedded in “theory application” as theories on MLG are employed to analyse the structural, regional, and policy factors, which facilitate MLG between Amsterdam and the EU.

Moreover, as the MLG framework circumambient to Airbnb has received little attention from scholars, there is a large quantity of new information that can be uncovered. As a result, this research takes an explorative approach. The qualitative research method is the best match for this research topic as it allows for material to be gathered and then used to explore an evolving topic that requires a detailed scrutiny for deeper meanings (Babbie, 2013). By employing a qualitative research design, this thesis seeks to strengthen our apprehension of social processes and activities that are occurring between Amsterdam, the Netherlands, and the EU. Quantifying this relationship would not be suitable because it would not allow an in-depth understating of the phenomena.

3.2 - Setting Scoping Conditions and Case Selection Strategy

The sharing economy encompasses a vast range of platforms that offers a variety of goods and services. Therefore, to limit the scope of this thesis, the study focuses on the role of Airbnb in Amsterdam and the EU between 2016 and 2020. Although there are larger sharing platforms, such as Uber, Airbnb has a particularly contentious relationship with the EU and various other Member State cities, as explained in section 2.6. Moreover, attempting to examine the sharing economy as a whole would not yield a fruitful in-depth study of MLG: many different cities have a vast range of active platforms and it would be challenging to account for all of these comprehensively in a Master’s thesis.

(39)

Amsterdam has been selected as a case study in this thesis for a variety of reasons. Firstly, because Amsterdam and the government of the Netherlands have an effective archiving system in both English and Dutch, thereby allowing for documents to be obtained in both languages. Although not a native Dutch speaker, the researcher of this thesis has a sufficient comprehension of Dutch (B1). Where documents were not able to be translated, a Dutch tutor at the Vrije Universiteit Amsterdam was consulted. Secondly, it would be extremely difficult to analyse the relationships of all European cities within the EU and with the SE. Indeed, if such an approach was undertaken, the quality of the thesis would be undermined as the structural, relational, and policy factors could not be accounted for in a meaningful way. In addition, Amsterdam has been chosen as it is hoped by picking a city that has responded to the onset of Airbnb by regulating it rather than banning it (such as Barcelona) will offer substance to the discussion situated around MLG. Finally, as the SE is a relatively new phenomenon, few local governments in Europe have sufficient documentation on the topic of Airbnb in their city archives that would allow for a fruitful discussion.

3.3 - Data Collection: Document Analysis 3.3a – Search Strategy

To investigate the research question, a document analysis was undertaken to collect the data for this thesis. Once the research topic was established, the next task was to develop criteria by which the literature would be assessed based upon the inclusion/exclusion criteria, key search terms to be used in the research, and the time span. The search in the relevant literature was taken between January 2016 and September 2020 and retrieved evidence appraised. The initial search generated papers from online databases from the local (City of Amsterdam), National (Government of Netherlands), and European levels. In total there were over 17356 documents from the European Commission website. To further narrow down the number of

(40)

the documents a keyword search was undertaken to reduce the number of papers. The search included the following key words:

• Sharing economy • Collaborative Economy • Airbnb

• Europe • Deeleconomie

These lists were downloaded and scanned by title and abstract for their importance and relevance to the sharing economy, the collaborative economy, and to the Netherlands. Those who did not meet the scope of the review were excluded. For those that did, a full copy was obtained. An in/out form was constructed based on the inclusion criteria below. Hand searching of the reference lists of included and relevant papers was conducted, and references from recent relevant systematic reviews were followed up. These documents were analysed and extracts used for the analysis are provided on a code a sheet that will be attached to the Appendix. In total 34 documents were identified that fitted inclusion criteria.

3.3b – Document Sampling

Documents can be primary or secondary data (Gross, 2018). For the thesis, primary data will be examined as they supplied a first-hand account about an occurrence or event, which are not accompanied, by an interpretation or an analysis (Gross, 2018). For the preliminary search, the databases and resources below were searched.

The following electronic databases were searched:

• The City of Amsterdam, News archive (Nieuwsarchief)

(41)

• Publications Office of the European Union

• The European Economic and Social Committee Publications

• The European Urban Knowledge Network

• EUR-Lex: EU Law

• EU Publications

• Government of the Netherlands (Rijksoverheid)

• Eurocities

• European Commission press room

• European Parliament press room

• Council of the European Union

3.3c - Inclusion and Exclusion Criteria

When considering the vast amount of available documents related to this research topic, it is vital that a set of clearly established inclusionary and exclusionary criteria is defined (Gross, 2018). In doing this, the documents selected are focused on the topic as well as on improving the authenticity and representativeness of those documents identified for the sample (Gross, 2018). Determining an inclusionary criterion is essential for providing a systematic document selection and for decreasing the number of irrelevant hits (Gross, 2018). On the other hand, an exclusionary criterion is significant as it helps limit the number of potential documents for the final sample (Gross, 2018).

The following criteria that has guided the selection of articles to be included in this thesis is as follows:

How Digital Disruptors Influence Multi-Level Governance on the EU: The Case of Airbnb and Amsterdam

Faculty of Governance & Global Affairs