The collaborative economy at the start of a transition : a trade-off between trust and convenience

1

The Collaborative Economy at the Start of a Transition:

A Trade-off Between Trust and Convenience

Master Thesis | MSc Business Administration | Marketing Track 01-07-2016

Author: Timo van de Laar | 11083700 University of Amsterdam Supervisor: Nicole Stofberg

2

NOTE: This research is part of an overarching research project in collaboration with ShareNL. ShareNL is a knowledge platform in the Collaborative Economy here in the Netherlands. They have close ties with all the major Dutch sharing platforms and they actively engage in research for the Dutch government. ShareNL investigates consumer behaviours, attitudes and motivations across various platforms in the Netherlands, as well as the barriers that prevent consumers from joining the new economy (http://www.sharenl.nl/#sharenl).

3

Statement of originality

This document is written by Student Timo van de Laar who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

4

Table of content

1.4 Relevance and Contributions 14

1.5 Thesis structure 15

2. Literature Review 16

2.1 Collaborative Economy 16

2.2 Organisational forms, trade forms and sharing sectors 18

2.3 Type of participation 22

2.4 Shift from P2P to P2B2P 24

2.6 Role of trust 28

2.6.1 Interpersonal Trust 29

2.6.2 Institutional Trust 30

2.6.3 Trust and active participation behavior 33

2.6.4 From trust to emotional attachment 35

2.7 Attachment theory 36

2.8 Conceptual model 40

3. Methods 42

3.1 Research design, Data collection and Sample 42

3.2 Measurement of variables 44

3.2.1 Translation process 44

3.2.2 Independent variable: personal contact 45

3.2.3 Mediating variables: interpersonal and institutional trust 45

3.2.4 Moderating variable: emotional attachment 45

3.2.5 Dependent variable: active participation behavior 45

3.2.6 Control variables 46

3.3 Data cleaning and Case Selection 46

3.4 Sample 48

5 4. Results 51 4.1 Descriptive statistics 51 4.2 Pre-analysis 52 4.3 Hypotheses testing 57 4.3.1 One-Way Anova 57

4.3.2 Process Model 4: Mediation effect 58

4.3.3 Process model 14: Moderated Mediation 61

5. Discussion 68

5.1 Findings 68

5.2 Contributions and practical implications 71

5.3 Limitations 73

5.4 Future research 74

6. Conclusion 75

7. References 77

Appendix A: Questionnaire active participants 85

Appendix B: Recoding of Variables 86

Tables

Table 1. Overview CE 19

Table 2. Willingness to participate per sector (ShareNL, 2015) 22

Table 3. Overview Hypotheses 39

Table 4. Frequencies and percentages Platforms 47

Table 5. Frequencies and percentages of all respondents in the Study (N=547) 48

Table 6. Descriptive Statistics (N=483) 52

6

Table 8. Correlation Matrix: Means, Standard Deviations, Correlations (N=483) 56

Table 9. One-Way Anova Hypotheses 1 57

Table 10. Process Model 4: Mediation role interpersonal trust 58 Table 11. Process Model 4: Mediation role institutional trust 59 Table 12. Process Model 14: Moderation effect interpersonal trust 62 Table 13. Process Model 14: Moderation effect institutional trust 64

Table 14. Overview Accepted/Rejected Hypotheses 66

Figures

Figure 1. Conceptual Model 41

Figure 2. Overview conceptual Model Mediation with effects 60

7

Abstract

The sharing economy is a new economy with huge potential. To reach this potential both the providing side and the using side should equally take part in it and remain participating actively over time... There is a shift taking place from ‘more personal’ peer-to-peer platforms to ‘more convenient’ and ‘impersonal’ peer-to-business-to-peer platforms, which could result in lower trust and subsequently lower active participation, especially on the provider side they bear most of the risks of sharing. Lower participation on the supplier side could be detrimental to the sharing economy’s sustainability as it would create a misbalance between supply and demand and therefore warrants scholarly attention. Moreover, previous research often focuses solely on users, but never considers the motivations and behaviors of providers. There the main research question is as follows: To what extent does personal interaction influence active sharing behaviors and what is

the role of trust herein? This is examined through a large cross-sectional survey, which is send to

the active participants of the Collaborative Economy in the Netherlands (n=555)

Analysis of the results showed that there was no significant relation between trust and active participation behavior. However, the results did find personal contact to lead to higher levels of trust, and significantly higher levels of active participation behaviours of providers. Since the success of the sharing economy rests on securing the participation of providers this finding demonstrates that the current shift towards ‘convenience driven’ and ‘impersonal’ P2B2P platforms is problematic. Contrary to expectations, this mechanism however is not explained through trust and warrants further attention. Also, emotional attachment had a positive relation with active participation behavior, indicating that providers want to share products to which they are attached more often, than products to which they are not attached. Future research should further examine the effects of emotional attachment. Also it should try to get insights in more

8

sectors and platforms to allow comparison, as for this research mainly Peerby participants were used.

9

1. Introduction

On February 2, 2015 Amsterdam was declared as first Sharing City of Europe, which makes Amsterdam the precursor of the so called Collaborative Economy (Sprang, 2015). According to Kajsa Ollongren, alderman of economic affairs Amsterdam, the Collaborative Economy is an enormous opportunity for Amsterdam. As the first sharing city of Europe, Amsterdam wants to make use of the chances that arise from the Collaborative Economy, in areas such as sustainability, social cohesion and economy.

Sharing is the most universal form of human economic behavior and as old as human kind (Belk, 2010; 2014; Botsman, 2010). We are born and bred to share and collaborate, yet the Collaborative Economy is born with the internet age. The Collaborative Economy (CE), also known as sharing economy, access-based consumption or collaborative consumption, is a new mega trend that is growing rapidly (Hartl et al, 2015).

It is a new type of business in which consumers borrow and rent out their personal belongings, which they are not using at the time, to complete strangers on the internet (Frenken and Meelen, 2014). Often there is a user review system involved, which is of great importance in order to generate a sort of mutual trust between the consumers and make the CE work (Botsman and Roger, 2010). Due to the internet the original middleman is removed, which enables consumers to share and collaborate with distant and previous unrelated strangers in new ways and on a much larger scale (Belk, 2014; Botsman, 2010). As a result people are sharing everything now: each other’s car, homes, equipment, books, clothing’s and so on (Baneke & Harmanides, 2015). Yet whilst the internet can facilitate the sharing of virtually everything, sharing platforms are still struggling with the best business models with which to encourage and sustain these sharing activities.

10

The global CE revenue was estimated at $15 billion in 2013, with the potential revenue of $335 billion in 2025 (PwC, 2014). According to Botsman (2010) the CE will be as big as the industrial revolution and mean the end of hyper consumption. The old wisdom of ‘you are what you own’ changed to a post-ownership economy with ‘you are what you can access’ (Belk, 2014). We do not want to own stuff anymore, but we want the needs it fulfills (Botsman, 2010).

Early sharing platforms encouraged the borrowing and lending of stuff amongst neighbours for free, like Couchsurfing and Peerby. But soon other sharing platforms also started to operate on for-profit models, such as Airbnb and Blablacar. This new type of business has already disrupted major industries, such as the hospitality- and transportation industry, because one gets convenient and cost effective access to resources without having to deal with financial, emotional, or social issues that come with ownership (Eckhard & Bardhi, 2015; Owyang, 2013). This shift to for-profit models is happening, because consumers are after utilitarian value, rather than fostering social relationships (Eckhard & Bardhi, 2015). In order to get competitive advantage, sharing platforms respond on consumers’ demand for lower costs and convenience, by acting as intermediator between both parties.

1.2 Research topic

According to Belk (2010) the more people who share, the richer the possibilities become. The success of the CE depends not only on demand, but also on supply. The intention to participate in the CE as a provider and share your goods with others is influenced by multiple factors, of which trust is the central element (Decrop & Graul, 2015). The forum of Young Global Leaders (2013) state that ‘trust is the social glue that enables collaborative consumption marketplaces and the sharing economy to function without friction’. In fact, potential providers see trust as the biggest

11

barrier to actively engage in the CE (Campbell Mithun, 2012). Thus, for the CE to work, it is important that the platforms create trust among their users.

A central tenant of this research is to try and answer whether convenient “for profit” business models are equipped to create the trust needed to sustain collaboration over time by providers or so called active participation behavior? In order to do so, this research posits that first of all researchers in the sharing field have to make a distinction between two types of trust: interpersonal trust and institutional trust. Whilst researchers like to stipulate the importance of trust, few have looked at the difference between the two different types of trust (Ter Huurne, 2015), let alone whether both are necessary to secure active participation behavior over time.

According to Ter Huurne (2015) both forms of trust are necessary to secure active participation behavior, yet no research to date looks at whether this also applies to hybrid forms of sharing platforms in which users do not meet each other and all the interactions are taken up by the middleman, the platform itself. This type of platform is referred to as the peer-to-business-to-peer variant by Van de Glind (2015) and distinguishes itself from earlier sharing models in that both parties of the interaction do not meet face to face anymore. This begs the question whether interpersonal trust is still necessary in this platform type or institutional trust suffices to secure cooperation under this new model.

Research among peer-to-peer models has shown that people are more motivated by social factors and want to interact with each other (Botsman & Rogers, 2010; ShareNL, 2015). It could be problematic for start-ups to only focus on convenience and efficiency, since it will make the CE less enjoyable and appealing for people who participate based on social reasons. On Airbnb participants stopped trusting each other after platform changes led to a loss in communitarian

12

values (Hendrickson et al., 2015). For current providers lack of trust might even mean that their number of transactions will decrease or that they will stop sharing (Ter Huurne et al., 2015).

However, for some sectors creating trust might not be as important as we think. According to Belk (2010) people are less willing to share things to which they have a higher emotional attachment. This means that when a product holds little value to the provider, than the risk of lending it will matter less. This could be an important reason why in some sectors there is a higher percentage of providers than in other sectors. Moreover, it will mean that for every sector a different level of trust is needed, depending on the level of emotional attachment that comes with the products being shared.

1.3 Research objective

The purpose of this paper is to get a deeper understanding of the influence of trust on the sharing behavior of providers in the Collaborative Economy. According to the literature there are two forms of trust, institutional- and interpersonal trust, that are of great importance for active participation behavior. Whereby peer-to-peer (P2P platforms create both types of trust, peer-to-business-to-peer platforms (P2B2P) mostly just lead to institutional trust, since consumers will only meet the intermediator. The question is: will institutional trust be enough to lead to active participation behavior of providers?

According to multiple studies institutional trust alone will not be enough (Ter Huurne et al., 2015; Celata et al., 2015). However, if these studies are right, how could we explain the success of the peer-to-business-to-peer platforms with a high retention rate of 44-60% (personal interview, ShareNL). This thesis asserts to understand this, we need to look at emotional attachment. The pre-assumption is that if emotional attachment is low, than institutional trust will be enough, but if emotional attachment is high, institutional trust will not be enough. If this assumption is proven

13

true, it will mean that platforms should only move from peer-to-peer to peer-to-business-to-peer in sectors where there is a low average of emotional attachment to the products.

This will be examined by comparing the importance of institutional- and interpersonal trust for providers on multiple P2P and P2B2P platforms in different sectors. The main focus will be on the moderating role of emotional attachment on the relation between providers’ trust and active participation behavior.

To examine this the following research question is made:

1. To what extent does personal interaction influence active sharing behaviors and what is the role of trust herein

The following sub questions are made to make sure that the whole research objective is covered:

1.1 To what extent does the level of interpersonal and institutional trust for providers depend on the level of contact that providers experience with the user of their product?

1.2 To what extent does interpersonal and institutional trust have an influence on active participation behavior?

1.3 To what extent does emotional attachment moderate the relationship between institutional- and interpersonal trust, and active participation behavior of providers?

In order to answer these research questions different hypotheses will be developed and tested, whereby all aspects will be taken in consideration.

14 1.4 Relevance and Contributions

As earlier mentioned the Collaborative Economy is a very new economy and because of that there is still a lack of research in this field. Most research about the role of trust only focuses on potential participants (Decrop & Graul, 2015; Hamari et al., 2015; ), but does not take in consideration how trust works for actual participants. When research does look at the actual participants, they only look at the influence of trust on users, but none of them looks at the influence of trust on providers, even though providers are pivotal to make the CE work (Ter Huurne et al., 2015).

Also, they only look at the importance of trust in general, whereby they do not make distinction in how different factors of trust may differ in importance for each type of platform. Even though it might be possible that platforms require different forms of trust, based on the level of personal contact that they facilitate.

Previous research does not look at the influence that emotional attachment has on the importance of trust, which might be an important factor in order to know when trust is necessary. By shedding light on these topics, we will try to enrich the literature on the CE and get a deeper understanding on when and why which type of trust is important for providers.

To conclude, this research will have practical relevance, since acquiring new insights in the role of trust can lead to further development of the CE. Online platforms can use the information to improve their platform in a way that makes sure that the right form of trust is created, based on the level of personal contact that they facilitate. With this research we hope to enable start-ups within the CE to improve their positioning and develop their platforms in a way that optimally attracts the providers for their sector.

15 1.5 Thesis structure

This thesis is structured as follows. In the next chapter the literature is reviewed, hypotheses are outlined and the conceptual model is constructed. Starting with defining all variables and their role for active participation behavior, followed by the hypotheses, and ending with the conceptual model. Chapter 3 describes the research methodology and subsequently, chapter 4 presents the results. In chapter 6, the results are discussed and practical and theoretical implications are stipulated. Lastly, chapter 7 concludes the thesis.

16

2. Literature Review

This chapter starts by defining the Collaborative Economy. First its organisational forms, trade forms and sharing sectors will be outlined, to make clear which forms will be understood as part of the CE. Secondly, the two types of participation (provider and user) will be described, to give insights in how both side of the CE differ from each other. Subsequently, the shift from P2P to P2B2P, and its possible effect on the relation between personal contact and active participation behavior, will be explained. Then, a distinction between interpersonal trust and institutional trust is made, and their important mediating role in the relation between personal contact and active participation behavior is described. Next, the potential moderating role of emotional attachment on the relation between trust and active participation behavior is outlined. This chapter concludes by giving an overview of the hypotheses that are constructed throughout the chapter, and implementing this in a conceptual model.

2.1 Collaborative Economy

The term Collaborative Economy (CE), is used for a broad amount of ‘economic systems of

decentral networks and marketplaces which make use of the value of under-used goods and services by bringing supply and demand directly together, so that traditional institutional in-between parties become unnecessary’ (ShareNL, 2015). Yet, Frenken and Meelen (2014) argue

that there are more and more companies who, unfairly, claim to be part of the CE to lift on the positive and progressive connotation that comes with it.

According to Frenken and Meelen (2014) many of the platforms that are considered as part of the CE have forms of ‘semi-sharing’. They define the CE in a narrow sense as: “Consumers granting

17

money” (Bocker & Meelen, 2015). This definition will be used for in this research, which excludes

‘semi-sharing’ practices from the scope of the Collaborative Economy.

So-called ‘semi-sharing’ occurs when people connected to a certain sharing platform are not making use of their idle capacity (Frenken & Meelen, 2014). For example, UberPop is a platform whereby anyone can sign up as a taxi driver, to bring another person from A to B with their own car, without the obligation of having a taxi license. When, with UberPop, a driver takes passengers on a specific trip that (s)he was going to make anyway, (s)he is using his/her idle capacity, since his/her otherwise empty extra seats are now being used by others. This form will be considered as part of the CE. In case the driver is just making the trip to bring another person from A to B, it concerns a taxi service. This latter form is seen as ‘semi-sharing’, which we will not consider as part of the CE.

Airbnb is another example of a platform where there are cases of ‘semi-sharing’ that are misinterpreted as part of the CE (Frenken & Meelen, 2014). When people are permanently living in another house, and meanwhile renting their own house on a continuous base on Airbnb, it means that they are in fact running an illegal hotel. This will not be considered as part of the CE. Only when people rent their house while they are temporary somewhere else (e.g. vacation, business trip or family visit), they are utilizing their idle capacity, since their house would otherwise be uninhabited.

Besides ‘semi-sharing’ the CE is often confused with Business-to-Peer (B2P) models. An example of a B2P model is Zipcar (Bardhi & Eckhardt, 2012). Zipcar is a car sharing service, whereby consumers get temporary access to cars owned by a company, in return for a membership fee. In this case we speak of access based consumption. Bardhi and Eckhardt (2012) define the access based consumption as: ‘Transactions that may be market mediated where no transfer of ownership

18

takes place, whereby the consumer is acquiring consumption time with the item, and, in market mediated cases of access, is willing to pay a price premium for use of that object’. Just like with

the CE, consumers get temporary access to products or services, so that they do not have to own them. However, just like ‘semi-sharing’, it is different from the CE in that it does not concern the sharing of one consumer’s idle capacity with another consumer. The providing party is a company, and therefore this will not be considered as part of the CE.

Thus, for the purpose of this research we distinguish the CE in a narrow sense, which means that it concerns the sharing of under-utilized physical assets between consumers. The main reason why we have chosen for the narrow definition is because the goal of this research is to examine the effect that product attachment has on provider’s active participation behavior. For ‘semi-sharing’ practices emotional attachment will not be relevant, because people are running a sort of business by sharing the products on a continuous base. Since they are not using the product themselves they probably will not feel attached to it either. In case of B2P there is no providing consumer who can feel attached to the product, because the providing party is a company. In short, in both cases emotional attachment to the product is irrelevant and thus its effect on active participation behavior cannot be measured.

2.2 Organisational forms, trade forms and sharing sectors

According to ShareNL (2015, 2016) the CE exists of multiple trade forms, sharing sectors and organisational forms (table 1). However, their broad definition differs from our narrow definition on several aspects. Therefore, we will now outline which forms and sectors of their broad definition are also part of our narrow definition (table 1), which will then be explained in further detail.

First, trade forms can be: buy, hire, lend, give, trade and share. Thereby we will only focus on hire and lend, since these forms are temporary. Secondly, the sharing sectors can be divided into: goods,

19

housing, transportation, logistics, energy, money, food, healthcare, knowledge and services. For this research all types of sharing sectors will be considered, because it does not matter what type of goods and services are being shared. At last, organisational forms can be peer-to-peer (P2P), peer-to-business-to-peer (P2B2P), business-to-peer (B2P), and business-to-consumer (B2C). As explained earlier, only the organizational forms P2P and P2B2P will be considered for this research.

Table 1. Overview CE

Broad definition (ShareNL) Narrow definition (this research)

Trade forms Buy

Hire Lend Give Trade Share Hire Lend Sectors Goods Housing Transportation Logistics Energy Money Food Healthcare Knowledge Services Goods Housing Transportation Logistics Energy Money Food Healthcare Knowledge Services

20 Organizational forms Peer-to-peer

Business-to-business Cooperative Peer-to-business-to-peer Peer-to-peer Peer-to-business-to-peer Sectors

The CE operates on both the supply and demand side of the economy. Everything is exchanged: from power drills, to knowledge, to single meals, to car rides, and so on. All these products and services can be divided over different sectors. At this moment there are ten recognised sectors in which the CE operates (ShareNL, 2015)(table. 1). The three most important and more evolved sectors (goods, housing and transportation) will be described in further detail, to get a better understanding about the type of platforms that these sectors contain.

The goods sector concerns the sharing of goods; that can be anything, from power drill, to partytent, to beamer (ShareNL, 2015). Peerby is a well-known example of a platform in the Netherlands where all types of goods are exchanged. It is already active in both organizational forms, with Peerby as P2P platform and Peerby GO as P2B2P platform. On Peerby people can borrow goods from their neighbours for free, whereby they will have direct contact with each other. Whereas on Peerby Go people can hire goods from their neighbours for small monetary compensations, whereby the platform will function as intermediator by picking up and delivering the goods (Peerby, 2015).

The housing sector concerns the sharing of accommodation (Airbnb), workspace (ShareDesk), parking spots (Mobypark) and gardens (Tuintjedelen) (ShareNL, 2015). ‘Tuintjedelen’ is an example of a platform that connects people who are looking for a garden, and do not have one

21

themselves, with gardeners who want to get out more of their garden by sharing it with other gardeners.

The transportation sector is about car sharing (Snappcar), ride sharing (Blablacar) and bike sharing (Cycleswap) (ShareNL, 2015). For example Cycleswap is a platform on which individuals can rent bicycles from each other (Knobben, 2016). It recently has been bought by the American company Spinlister, which is a platform that also provides the sharing of surf- and snow gear.

Organisational Forms

Traditional research only focused on the older organisational forms of the CE, which concerned P2P platforms. Beside P2P platforms, this research will also look at a newer, rapidly rising, form, called P2B2P. Not only are new platforms starting with P2B2P forms, also P2P platforms are shifting to, or adding, P2B2P platforms. The difference between the two organisational forms will now be explained. Later on we will explain what this shift from P2P to P2B2P will mean for the active participation behavior of providers.

1. Peer-to-Peer (P2P)

The Peer-to-Peer form can be described as marketplaces that eliminate the need for the traditional middleman and bring one person in direct contact with another person, so that demand and supply directly come together (ShareNL, 2015). Thus, on these platforms the user can directly lend or hire stuff from the provider, which he/she is temporary not using. Hereby, the provider and user will have personal contact to fulfil the transaction.

2. Peer-to-Business-to-Peer (P2B2P)

With Peer-to-Business-to-Peer forms there is an intermediary in between the two parties (demand and supply), to facilitate the hiring or lending of the good (ShareNL, 2015). Both user and provider

22

are still individuals, but the difference with P2P is that they do not have to meet each other in person, due to the intermediary.

2.3 Type of participation

Within the CE there are two kinds of participants: providers, who lend their own products and services to others, and users, who borrow other people’s products and services. People can either participate as provider, user or both. According to van de Glind (2013) more potential consumers were willing to participate as a user (43,4%), than as a provider (31.9%). Additionally, ShareNL (2015) shows that there is little difference in willingness to participate as a user or a provider in the goods sector, whereas in the housing and transportation sector this difference is much greater (table 2).

Table 2. Willingness to participate per sector (ShareNL, 2015)

User Provider

Housing 58% 13%

Transportation 37,50% 24,60%

Goods 52,20% 50,70%

These differences are problematic, because for the CE to operate successfully, it is necessary that both supply and demand are high (Belk, 2010). Especially the willingness to participate as provider is problematic, since this varies much more per sector, than the willingness to participate as user. This bares the question why as a provider some sectors are less attractive than other sectors, whereas for users there is not such a significant difference?

Unfortunately, previous research only looks at the user side (Bardhi & Eckhardt, 2012; 2015; Möhlmann, 2015; Ter Huurne et al., 2015), resulting in that there is little known about the provider

23

side of the CE. Due to the great difference in willingness to participate as a provider, this research will only take the provider side into account. As possible reasons for these differences we will look at the factors trust and emotional attachment, which we will discuss later in further detail.

Active participation behavior

As mentioned earlier, previous research only has focused on people’s intentions to share, which does not measure the active participation behavior. The problem with people’s sharing intentions is that it is not always converted into real action (Hamari et al. 2015). In order for the CE to reach its full potential it is necessary that both supply (providers) and demand (users) are high (Belk, 2010). Therefore, the willingness to participate as provider should increase, and these intentions should be converted into active participation behavior. To investigate usage patterns more accurately it is necessary to measure the actual usage. Therefore, this research will not look at intentions, instead it measures active participation behavior of users and providers in the CE, with a main focus on providers.

According to Hamari et al. (2015), intentions to participate in the CE depend on the person’s attitude, the perceived economic benefits of the CE, the perceived sustainability of the CE, and the perceived enjoyment of the CE. Moreover, Decrop & Graul (2015) state that trust is the central element for a person’s intention to participate in the CE.

Unfortunately these intentions do not always end up in actual behavior, and thus active participation behavior. Active participation behavior is not about why or when someone will participate, or in other words his/her intentions. Instead, it is about how often a person makes use of a certain platform. This might vary largely over individual persons. Whereas one might lend his/her apartment once a year, another person might do it every month.

24

According to the Theory of Reasoned Action (TRA), intentions are the main antecedent of voluntary behavior, and thus is likely to influence future participation behavior (Ajzen and Fishbeien 1980). In other words, intentions to participate in the CE is the main antecedent of active participation behavior in the CE.

Hypotheses:

As showed by ShareNL (2015), there is higher willingness to participate as user, than as provider. Therefore, following the reasoning of TRA, it is expected that users will show higher levels of active participation behavior than providers.

H1: Users show higher levels of active participation behavior in the CE than providers.

2.4 Shift from P2P to P2B2P

The CE is undertaking a shift to higher convenience, which is according to Bardhi and Eckhardt (2012) a good thing. In fact, they suggest that the CE should be replaced by access based consumption, whereby social interaction becomes irrelevant. This is because their study about the motivations of Zipcar users showed that consumers ask for convenience and value for their money. Consumers do not care about social interaction with other consumers or the company, and are mistrustful of other consumers. They just want to use Zipcar in an anonymous way, as long as the company makes sure that the sharing system is controlled and fair. Consumers do not want the hassle that comes with the CE, they are just participating based on self-motivated reasons.

Zipcar is a B2P platform, which does not include consumers on the provider side. But, the shift to higher convenience also includes new platforms with a P2B2P form, such as PeerbyGo, whereby consumers also participate as the providing party. Just like Zipcar, participants of Peerby indicated

25

to care for higher convenience (Peerby, 2015). In response, Peerby launched Peerby GO, which arranges the pick-up and delivery of people’s stuff, to save time and energy (Peerby, 2015).

Convenience is according to van de Glind (2015) a key factor for the Collaborative Economy, and can be defined as the ease to obtain, use or reach, anything that saves or simplifies work. If it is true that consumers only ask for higher convenience, Airbnb’s rebranding, which highlights ‘people, places, love and community’, will be a misstep. As the main reason to use Airbnb, is to get value for their money, especially in expensive cities. Moreover, consumers want to have the whole place for themselves, so that they do not have to share it with others.

According to Daan Weddepohl (founder of Peerby), convenience and reciprocity are required to make peer to peer lending workable on a daily base (Peerby, 2015). Inconvenience of P2P platforms is one of the main factors that limits the growth of these platforms. The time that a provider has to wait for the user of his or her good, car or apartment can be seen as inconvenient, and as a barrier to take part in the CE. Therefore, more and more marketplaces are providing intermediaries who facilitate the lending or renting of goods between a provider and a user.

Although the shift from P2P to P2B2P, in order to create higher convenience, sounds like the ideal option, it can have problematic consequences. The first problem is that Bardhi and Eckhardt’s (2012; 2015) reasoning, wherein they only look at the user side of the CE, cannot be generalised to the whole CE. Their Zipcar example is a B2P platform, whereby consumers take only part on the user side, which excludes people who want to participate as providers. Barhi and Eckhardt do not consider that there are numerous studies (Bhattacharya et al., 2009; Botsman & Rogers, 2010; Frey & Jegen, 2001; Ikkala & Lampinen, 2015, Van de Glind, 2013) that show that people share for more than self-regarding motivations alone. Motivations are distributed unevenly between users and providers, and providers are even more socially motivated (Bocker & Meelen, 2015).

26

In 2013, Fastcoexist ran the headline: “People get in it for the money, but stay in it for love”. Whilst the cost-savings seemed a great incentive for people who had never participated, yet the majority of participants who had actually tried it, indicated to participate based on social motivations. Interestingly, this trend not only applies to platforms that deal in small transactions such as Peerby (Van de Glind, 2013), but has also been mentioned in relation to Airbnb. As Ikkala and Lampinen (2015) explain: “while the possibility for earning money is an important factor in igniting participation, the social aspects of network hospitality play a central role in sustaining hosts’ motivation to keep participating.” (p. 1034).

It appears that when participants actually help others and have nice social interactions with like-minded people, these platforms create unanticipated, ‘priceless’ value, which causes people to stay in it for the long haul and go the extra mile (Ikkala and Lampinen, 2015). None of which will happen when platforms become very impersonal, with organizational forms like B2P and P2B2P.

A second problem that emerges from the shift to P2B2P, is due to its impersonal character. Providers run all the risk of sharing and –thus run the risk of free riding- may prefer a system that is a little more effort but provides more safety of their property. After a bad experience people are less likely to reciprocate (Hoffman et al., 1998).

People are far more likely to mistreat goods if you remove the human element to ensure a higher level of convenience (Bhardi, 2014). There are already examples where the shift from P2P to P2B2P backfired, such as for Turo. Turo is P2P car sharing platform that partnered up with General Motors Onstar division to give users the ability to unlock car’s with their cell phones, so that users do not have to exchange keys anymore (RelayRides, 2013; Van de Glind, 2015). However, this change had a negative effect on accountability and let to more conflicts between users, due to that

27

users did not have to meet face to face anymore to exchange keys. Because of this, Turo decided to go back to the old system.

In short, P2P enables individuals to get to know each other through personal contact, which has two major benefits: (1) It is more likely to trigger social motivations, which on the long term leads to active participation behavior; (2) Personalization generates trust and fosters an environment in which people take care of the greater good rather than their own narrow self-interests, resulting in more positive experiences, which in turn keeps people motivated and active in the CE. This latter argument will be further explained in the next paragraph, where we will go into detail about the role of trust on providers active participation behavior.

Unfortunately, only few P2B2P platforms exist so far, which makes it hard to acquire a sample that is evenly distrusted over the two platform types and be able to make a reliable comparison. To solve this problem we look at the level of personal contact that a provider experiences, as an alternative for platform type. Whereby personal contact below average will be considered as P2B2P, and personal contact above average will be considered as P2P.

Hypotheses

As already explained it is assumed that P2P platforms trigger social motivations, which on the long term lead to higher active participation behavior. Therefore we assume that the shift from P2P to P2B2P, although leading to higher levels of convenience, will result in lower active participation behavior of providers.

H2: There is a positive relation between personal contact and active participation behavior, so that higher levels of personal contact will result in higher levels of active participation behavior.

28 2.6 Role of trust

An important explanation for the link between provider active participation behavior and personal contact is trust (Pavlou & Gefen, 2004). Trust plays an important role when there is a situation with risk, uncertainty and interdependence (McKnight & Chevany, 2002; Pavlou & Gefen, 2004). In the CE participants have to deal with different kind of trust issues, such as opportunists, damage to personal belongings, endangerment of personal safety and liability issues (Ter Huurne et al., 2015). Both provider and user show some kind of vulnerability to the other, and have certain expectations on how the other is going to behave.

Reasons that withhold people from participating in the CE are mainly based on issues with trust in the other party. Others may not treat your stuff with the same level of care, which can be very annoying (Brunning, 2015). Moreover, people might break your stuff, which will probably have a negative effect on your trust in strangers in future sharing of your goods. These barriers are especially relevant for providers in the CE, since providers have to trust strangers (users) with their personal belongings. When you lend out your personal belonging you are running most of the risks and are more vulnerable to exploitation.

To define trust, we will use the same definition as being used by Mayer et al. (1995):

‘The willingness of one party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party’.

Research to date mentions the key role that trust plays in the CE (Belk, 2014; Botsman & Rogers, 2010; Hartl et al., 2015). Moreover, the results from an online survey conducted by PwC (2014) show that first condition for participating is trust, whereby 89 percent of participants agree that the

29

CE is based on trust between providers and users. If trust is such an important feature for consumers to participate in the CE, it is necessary to know how this trust can be generated.

The current shift from P2P to P2B2P will also cause a shift in the type of trust that the platforms produce. Whereas on P2P platforms consumers can built trust through personal contact with one another, on P2B2P platforms they will have to trust the intermediator, as there will be less or no contact between the provider and user.

In the literature three types of trust are identified: (1) dispositional trust, (2) interpersonal trust, and (3) institutional trust (Decrop & Graul, 2015; Ter Huurne et al., 2015). The first, dispositional trust, will be excluded from the scope of this study, since it describes an individual’s general attitude of seeking trustworthiness, for which it does not matter if a platform has a P2P form or a P2B2P form. Interpersonal trust is the trust that a consumer has in another consumer on a personal level in a specific context. This means that in some situations consumers might trust others more than in other situations. Institutional trust concerns the reliance on the system or institution that allows the sharing as such. In case of the CE it considers trust in platforms that function as intermediaries, such as Airbnb (housing) or PeerbyGo (goods).

2.6.1 Interpersonal Trust

The key factor in building trust between the provider and the user is interaction between the two parties (Lu et al, 2010). Interpersonal trust is built through interactions between peers, because both sides of the transaction get familiar with each other, which might take away initial disbeliefs between participants (ShareNL, 2015). The participants of Peerby indicated that brief interactions with their neighbours when exchanging items with each other, increased participants’ general trust

30

in one another. Moreover, this getting to know each other is experienced as pleasant, leads to higher positive expectations about one’s neighbourhood and even resulted in spontaneous street parties. Interpersonal trust concerns the orientation of one actor towards a specific person (Simpson, 2007). It describes the trust that one person has in another person on a personal level in a specific context (Decrop and Graul, 2015). Interpersonal trust can be seen as the outcome of an actors expectations that a specific person will perform a certain action (Simpson, 2007). It occurs when individuals believe that the other individual is committed, has benevolent intentions and pro-social motivations. Interpersonal trust is based on three factors, namely: interaction between actors, reputation and perceived similarity (Ter Huurne et al. 2015).

2.6.2 Institutional Trust

Institutional trust is based on guarantees and recommendations of third parties, also known as intermediaries (Pavlou & Gefen, 2004). The intermediary gives the rules and creates the institutional framework on which the market operates. A person will trust the in-between party when it for example it offers guarantees through safety policies (Ter Huurne et al., 2015).

Pavlou and Gefen (2004) provide insights in how institutional trust is built for consumers on online auction marketplaces. They state that institutional trust for online auction marketplaces is based on four aspects: (1) feedback mechanism, (2) escrow services, (3) credit card guarantees, and (4) trust in the marketplace’s intermediary.

However, compared to auction marketplaces, these four institution based structures will work differently for the CE. Whereas on online auction marketplaces the demand side is more at risk of suffering a loss when pursuing transactions (Pavlou & Gefen, 2004), in the CE the supply side is more at risk. This is because when transactions include permanent purchases, such as for online

31

auction marketplaces, buyers have a higher perceived risk, due to sellers potentially opportunistic behavior (e.g. not delivering the right product at the right time promised). Hereby the seller does not care about how the buyer will treat his/her sold product, since the seller will not use the product again. But, when the transaction includes a temporary exchange, such as for the CE, the supply side is at higher risk of opportunistic behavior. This is because the providing party is at risk that the user will damage the product, which needs to be returned to provider after the temporary usage, whereas the user only has to use the product for a short period of time.

For institutional trust we will only look at the fourth feature, known as ‘trust in intermediary’. Trust cues such as ‘reputation systems’ (feedback systems) and ‘credit card guarantees’ will not lead to different results between P2P and P2B2P platforms, because they are applicable for both P2P and P2B2P platforms, and will generate both institutional trust and interpersonal trust.

Trust in intermediary

An online intermediary is a third-party institution who facilitates transactions between users and providers by using the internet infrastructure to collect, process, and distribute information (Pavlou & Gefen, 2004). One of their main roles is to build trust and reduce risk. This can be done through the establishment of regulations, which restricts a user’s or provider’s ability to engage in opportunistic behavior. Moreover, they can provide guidelines of acceptable transaction behavior.

To define trust in intermediary the same definition as Pavlou and Gefen (2004) will be used:

The subjective belief with which a buyer believes that an intermediary will institute and enforce fair rules, procedures, and outcomes in its marketplaces competently, reliably, and with integrity, and if necessary, will provide resource for buyers to deal with seller opportunistic behavior.

32

Pavlou and Gefen (2004) state that intermediaries can reduce uncertainties in four ways: (1) Providing a reliable and secure environment, (2) establishing fair and open rules and procedures, (3) sanctioning, evaluating, or removing problematic sellers, and (4) encouraging benevolent transaction norms. Such activities generally build trust since it indicates that there is fair treatment and outcome. Providers who trust the intermediator should also trust the users, because of the perceived association with intermediary.

Hypotheses:

Interpersonal trust and institutional trust differ to each other in that interpersonal trust is a person’s trust in the other party, whereas institutional trust is a person’s trust in the facilitating party, or in other words the intermediary. On P2P platforms provider and user need to have personal contact with each other to make the transaction happen. Interpersonal contact is the key factor for building interpersonal trust, and thus we expect P2P to mainly lead to higher levels of interpersonal trust, and to a lower extent to trust in the intermediary. In contrast, on P2B2P providers will only have contact with the intermediary, whom from there handles the transaction. Providers will not have direct contact with the using party, and therefore have to trust the intermediary to assure that the using party will use the product accordingly. Therefore, we expect P2B2P platforms to mainly lead to higher levels of institutional trust, and to a lower extent of interpersonal trust.

H3a: There is a positive relation between personal contact and interpersonal trust, in that higher levels of personal contact will result in higher levels of interpersonal trust.

H3b: There is a positive relation between personal contact and institution trust, in that higher levels of personal contact will result in higher levels of institutional trust, but the relation is less strong than the positive relation between personal contact and interpersonal trust.

33 2.6.3 Trust and active participation behavior

The primary duty of trust is to reduce the expectations of opportunistic behavior and diminishes risk perception (Pavlou & Gefen, 2004). But ultimately, it will, together with perceived risk, have an impact on a providers intention to participate. This is based on theory of reasoned action, which states that expectations about the outcome of a given behavior will influence intentions to undertake that behavior (Ajzen & Fishbein, 1980). Hence, trust and perceived risk can be seen as beliefs that result in behavior intentions.

In first instance trust is used as a major social uncertainty reduction mechanism, that allows a person to subjectively rule out many undesirable possible behaviors on the party they trust and to bring all these possible outcomes to a more manageable level (Pavlou & Gefen, 2004). Once this is done, the provider will rationally assess if he/she wants to continue in sharing with the other party. Hereby, the behavioural intentions are the primary antecedent of voluntary behavior, which likely will influence future active participation behavior. Thus, intentions to share will over time translate into active participation behavior.

This indicates that institutional trust and interpersonal trust function as mediators in the positive relation between level of personal contact and active participation behavior. The question now is if trust in intermediary will generate sufficient trust to eventually translate into active participation behavior, and thus will have the ability to replace interpersonal trust (P2P)? Decrop and Graul (2015) state that interpersonal trust and institutional trust are both of crucial importance in order to facilitate an increase in providers’ participation intention on P2P platforms. Moreover, they suggest that the shift from P2P to P2B2P will have an impact on trust and active participation behavior. At first they thought that it would increase users institutional trust, but decrease users interpersonal trust. However, the results showed something different, in that the shift from P2P to P2B2P leads

34

to higher institutional trust and higher interpersonal trust for potential users. This is because users who trust an intermediator (institution), are also likely to trust the provider, since the user will associate the provider with the intermediator. Thus for users interaction with the intermediator will function the same as if they are interacting with the provider, and therefore still leads to higher interpersonal trust.

Hypotheses:

If the findings of Decrop and Graul (2015) about potential users are also applicable for providers it would suggest that P2B2P platforms would generate both types of trust (interpersonal and institutional), and therefore the shift from P2P to P2B2P will not result in less active participation behavior of these providers. Nevertheless, in our opinion the shift will have a different effect on providers trust and active participation behavior, than on users trust and active participation behavior. This is because interpersonal trust is of greater importance for providers, and will be built through different criteria. Moreover, multiple studies (Lu et al, 2010; ShareNL, 2015; Ter Huurne et al. 2015) state that personal contact is a prerequisite in order to build interpersonal trust. Therefore, based on the importance of interpersonal trust, we assume that providers are showing higher active participation behavior on P2P platforms, than on P2B2P platforms. This means that that mediating role of interpersonal trust in the relation between personal contact and active participation behavior is more significant than the mediating role of institutional trust. Or in other words, institutional trust is too a less extent an indicator of active participation behavior.

H4a: Personal contact indirectly affects active participation behavior, via interpersonal trust.

H4b: Personal contact indirectly affects active participation behavior, via institutional trust.

35 2.6.4 From trust to emotional attachment

Ter Huurne et al. (2015) state that the importance of the transaction, in other words the product being shared, results in different levels of trust. For example the sharing of an apartment requires higher levels of trust than sharing a power drill, due to its higher value. Also it might differ per person: for one person the sharing of a car has high perceived risks, since the car has high value and is his/her most important transportation, whereas for someone else sharing a car has low perceived risks, because he/she does not use the car often.

There are two moderating factors that will influence the importance of trust and active participation behavior, namely familiarity and perceived risk (Ter Huurne et al., 2015). Familiarity concerns the previous knowledge that one has about e-commerce and the particular platform, which can make that person more comfortable in trusting the other party (Li et al., 2007; Sutanonpaiboon and Abuhamdieh, 2008). When someone has positive previous experience with P2P and P2B2P platforms, he/she is more likely to trust the platform. However, the focus of this research is on the impact of personal contact, therefore familiarity will not be considered as relevant for this research.

Perceived risk on the other hand is of critical importance, as it has an impact on actual purchase decisions and thus active participation behavior (Kim et al., 2008; Xu et al., 2010). As earlier mentioned, risks in the Collaborative Economy can be (1) damage to the products that are being shared and (2) personal safety issues when the transaction happens at the providers home. The perceived risk is influenced by factors such as value of the product, the volume and the transaction costs (Li et al., 2012). The perceived risk will be higher when these factors increase.

We assume that perceived risk does not only depend on aspects such as financial value, but also depend on the influence of emotional value (emotional attachment) that products can have. Financial value has already been examined, for this research the focus will lie on the latter,

36

emotional attachment, which is expected to have a moderating influence on the relation between trust and active participation behavior and will be described in further detail in the next paragraph.

2.7 Attachment theory

The influence of perceived risk on trust is connected to the level of “attachment to products”. According to Belk (2010, p. 727): “people are more reluctant to share things to which they have a strong emotional attachment”. Hence it would seem that trust plays a much more critical role when “attachment to products” is high than when it is low. Or in other words, if a product holds little value to the provider, than the “dark side of lending” (i.e. risk that it is not properly looked after) will matter less. And thus when there is high emotional attachment, the perceived risk will be higher, and therefore the active participation behavior will be lower.

An attachment can be described as ‘an emotion-laden target-specific bond between a person and a specific object’ (Bowlby, 1979). The desire to make strong emotional attachments to particular others is a basic human need (Bowlby, 1980). Attachments vary in strength, whereby stronger attachments are associated with stronger feelings of connection, affection, love and passion.

According to Schifferstein & Zwartkruis-Pelgrim (2008) consumers can experience different levels of attachment towards the durable products that they own. To some products they stay attached and some others they easily dispose. They define emotional attachment as ‘the strength of the emotional

bond a consumer experience with a durable product’. When someone is attached to a specific

product, the object is special and will mean a lot to the person. Moreover, losing the product will feel as an emotional loss to the person. When people feel attached to products, they may even keep using them when they do not function properly anymore.

37

The most important aspect of emotional attachment is irreplaceability (Schifferstein & Zwartkruis-Pelgrim, 2008). Irreplaceable means that the product has a symbolic meaning to the owner that cannot be found in other products. Irreplaceability is based on a personal, idiosyncratic bond with the product, and thus the most important aspect of emotional attachment.

Important factors that affect the degree of attachment are memories and enjoyment (Schifferstein & Zwartkruis-Pelgrim, 2008). A person will be more emotionally attached to a product when this product evokes positive memories. These memories can be related to another person, a place or a specific event, and are the main driver of emotional attachment for older products. Enjoyment is a main driver of emotional attachment for newly acquired products, since people enjoy using these products. Enjoyment may become less important over time, because on average people get less enjoyment from products that are outdated. On the other hand memories will become a more important driver of emotional attachment over time, because objects that people have owned for a long time will elicit more memories, and will probably elicit even more memories over time.

The attachment theory by Bowlby (1979) says that ‘the degree of emotional attachment to an object

predicts the nature of an individual's interaction with the object’. Thus, individuals who are

strongly attached to an object (house, car, power drill, etc.) are more likely to be committed to, invest in, and make sacrifices for that object (Bowlby, 1980; Hazan & Shaver, 1994).

Moreover, the stronger one's attachment to an object, the more likely one is to maintain closeness to the object (Thomas, et al., 2005). If a person experiences stress from the external environment, (s)he often seeks for physical or psychological protection from the attachment object. Moreover, distress can occur when individuals experience real or threatened separation from the attachment

38

object. Thus when people have a high attachment to an object, they will try to protect that object and make sure they will not be separated from it.

Hypotheses:

We assume that providers with higher levels of emotional attachment will ask for higher levels of trust, in order to protect their product and make sure that their product is used properly. Moreover, based on the literature we suggest that especially creating interpersonal trust is important for providers, since they base their trust in the other party mainly on personal contact. Finally, we propose that when providers have very low emotional attachment to the product they do not care about if the user can be trusted, and therefore these people are showing higher levels of active participation behavior.

H5a: The indirect effect of personal contact via interpersonal trust to active participation behavior will be smaller when emotional attachment is higher.

H5b: The indirect effect of personal contact via institutional trust to active participation behavior will be smaller when emotional attachment is higher.

Table 3. Overview Hypotheses

Subject Hypotheses

Role H1. Users show higher levels of active participation behavior in the CE than providers.

Personal Contact H2: There is a positive relation between personal contact and active participation behavior, so that higher

39

levels of personal contact will result in higher levels of active participation behavior.

Trust H3a: There is a positive relation between personal contact and interpersonal trust, in that higher levels of personal contact will result in higher levels of

interpersonal trust.

H3b: There is a positive relation between personal contact and institution trust, in that higher levels of personal contact will result in higher levels of

institutional trust, but the relation is less strong than the positive relation between personal contact and

interpersonal trust.

Active participation behavior H4a: Personal contact indirectly affects active participation behavior, via interpersonal trust. H4b: Personal contact indirectly affects active participation behavior, via institutional trust.

Emotional Attachment. H5a: The indirect effect of personal contact via interpersonal trust to active participation behavior will be smaller when emotional attachment is higher. H5b: The indirect effect of personal contact via

institutional trust to active participation behavior will be smaller when emotional attachment is higher

40 2.8 Conceptual model

With this research we propose that the extent to which providers show higher levels of active sharing behavior depends on personal contact with the other party and the type of trust that is generated. To what extent interpersonal trust and institutional trust ultimately lead to providers active participation behavior is moderated by emotional attachment. The final goal is to get answer on our main research question: To what extent does personal interaction influence active sharing

behaviors and what is the role of trust herein?

The proposed hypotheses (table 3) and the relations between the variables of interest are visualised in the following conceptual model (fig. 1):

41 H1 H2 H3b H3a H4a H4b H5a H5b

Figure 1. Conceptual Model

Roles of variables

The variables of interest are personal contact (IV), interpersonal trust (MED), institutional trust

(MED) emotional attachment (MOD), and active participation behavior (DV). The independent

variable is personal contact (which is related to platform type). The dependent variable is active participation behavior, which indicates how often a provider shares his/her products. The relation between platform type and provider’s active participation behavior is mediated by trust. The mediator trust exist of interpersonal trust and institutional trust, which will both have a positive impact on provider’s active participation behavior. Finally, the relation between trust and provider’s active participation behavior is negatively moderated by emotional attachment, so that higher levels of emotional attachment will lead to lower levels of trust, and consequently to lower provider’s active participation behavior.

Personal Contact Active

Participation Behavior Interpersonal Trust Emotional Attachment 2.2 System Trust Institutional Trust

42

3. Methods

In this chapter the empirical process of this research is outlined. First the research design is described. Subsequently, the data collection technique is explained. Afterwards, it is described how the different variables of the conceptual model are constructed and tested. Then, the data cleaning and recoding for the sample is outlined, which are the first steps to make the data set ready for the analysis of the results. Next, an overview of the sample is provided. Finally, the statistical process will is provided, to explain how the survey data is analysed.

3.1 Research design, Data collection and Sample

This study is deductive and of explanatory nature, to test theory and establish causal relationships between the variables of interest (Saunders et al., 2009). The research will be quantitative, examined through a cross-sectional survey design. This means it will measure, through a questionnaire, when which type of trust is necessary for providers to lead to active participation behavior in the CE at a particular moment in time. A survey design is used to collect a large amount of data and allow comparison over these data (Saunders et al., 2009).

The survey questionnaire is constructed with the program SurveyMonkey, in collaboration with ShareNL, and consists questions from three different Master thesis’s, which are all part of the of the overarching research project of ShareNL and the University of Amsterdam. Consequently, the questionnaire also includes questions that are irrelevant for the purpose of this particular research. The goal of this research project is to acquire new insights in the behaviors, attitudes and motivations of potential- and active participants of the CE, in the Netherlands.

First a pre-test was done to test the constructed scales, and examine if all questions are understandable and rightly interpreted. For the pre-test a convenience sample was used, by using the networks of the three master student’s that are contributing to this research project (including

43

myself). The pre-test had a response rate of n=96, with a completion rate of 47% (n=45). With the pre-test the Cronbach’s Alpha’s of interpersonal trust and institutional trust were tested. Both scales were derived from Möhlmann (2015), consisting of three items each, with a Likert scale on a 7 point scale (totally disagree – totally agree).

The Interpersonal trust scale (Trust_i1, Trust_i2, Trust_i3) had bad reliability, with Cronbach’s Alpha =.584 (< .70). The corrected item-total correlation indicated that not all items have a good correlation with the total score of the scale (> .30). Trust_i1 had a corrected item-total correlation of .142 (< .30) and a Cronbach’s Alpha if Item Deleted = .823 (< .10). Therefore, Trust_i1 was deleted from the scale.

The Institutional trust scale (Trust_s1, Trust_s2, Trust_s3) had good reliability, with Cronbach’s Alpha =.785 (> .70). The corrected item-total correlation indicated that all items had a good correlation with the total score of the scale (> .30). However, Trust_s1 Cronbach’s Alpha if Item Deleted = .908 (< .10). Therefore, Trust_s1 is deleted from the scale.

Thus, the pre-test indicated that for both variables 1 item had to be deleted from the scale, which meant that the new scales of both interpersonal trust and institutional trust now consisted of only two items. However, it is preferable that these variables are measured by at least three items to have a better reliability (Guilford, 1952), therefore it was decided to use new items for the measurement of both variables. The new items were derived from Pavlou and Gefen (2004) and translated to Dutch. Later on this new variable will be explained in further detail.

Finally, the survey was optimised through the analysis of the scales of the variables and the feedback of the respondents. The respondents did not only fill in the survey, but also provided

44

feedback by indicating specific typos, certain errors in the questionnaire and by signalling questions that were difficult to understand (see appendix A for the final survey).

After optimising the questionnaire the survey was sent out. The population of the sample is people in the Netherlands, whereby a non-probability sample is used, so that only people who are active participating on the collaborating platforms have a chance of being selected. ShareNL used its partners to send the questionnaire by online means to active participants of all collaborating sharing platforms. All partners are online sharing platforms that are part of the CE in the Netherlands. For each partner the survey questions are modified in such a way that it corresponds with the sector in which it is active. It will be on a voluntary base through self-selection, whereby the mentioned platforms invite the sample members to fill in the survey. The survey was sent out and open on Friday 17th of June and closed on Monday 20th of June. Over a four day time period the survey was filled in by a total of 856 respondents, with a completion rate of 67% (n=555). Later on this sample will be further outlined.

3.2 Measurement of variables

Multiple items were constructed for each of the following variables: personal interaction,

interpersonal trust, institutional trust, emotional attachment, and active participation behavior.

3.2.1 Translation process

All items that are used in the questionnaire are either self-constructed or obtained from English studies. Since this research examines the Dutch population, whereby all respondents have to be Dutch citizens, all original items were translated from English to Dutch. Multiple third parties and experts were asked to second check the survey and newly added items, to make sure that all items were correctly translated from English to Dutch.