Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1
Innovation Studies Utrecht (ISU)
Working Paper Series
Different business models - different users?
Uncovering the motives and characteristics of B2C and P2P carsharing adopters
Karla Münzel
Laura Piscicelli
Wouter Boon
Koen Frenken
ISU Working Paper #18.01
2
Different business models - different users?
Uncovering the motives and characteristics of B2C and P2P carsharing adopters
Karla Münzela*, Laura Piscicellib, Wouter Boonc, Koen Frenkend
*Corresponding author
Innovation Studies Group, Copernicus Institute of Sustainable Development, Utrecht University
Princetonlaan 8a, 3584 CB Utrecht, The Netherlands
a [email protected] | +31 30 253 7301
b [email protected] | +31 30 253 6792
c [email protected] | +31 30 253 2708
d [email protected] | +31 30 253 6708
Prepared for Transportation Research Part D: Transport and Environment
Abstract
Carsharing is regarded to play an important part in the transition towards a more sustainable
mobility system by changing how cars are used and transportation needs are met. Over the past
decade, there has been considerable interest in understanding the characteristics and motives of
car sharers. Yet, studies have been mostly limited to small surveys in single cities. What is more,
past studies only covered traditional business-to-consumer (B2C) carsharing, ignoring the
3
growing popularity of peer-to-peer carsharing. The key question we pose in this study is whether
characteristics and motives differ between B2C and P2P carsharing users. We present survey
results among 1,836 Dutch citizens regarding the adoption, intention to adopt and non-adoption
of both B2C and P2P carsharing. We further look into the frequency of use and car purchase
avoidance among carsharing users. Finally, we investigate car owners who supply their car on
P2P platforms. We find that B2C and P2P carsharing adopters are very similar. The main
difference between the two users groups holds that B2C users are more frequent users with higher
income valuing the convenience of B2C carsharing. We conclude that as the convenience of P2P
carsharing is likely to increase with automatic locks and higher supply, user experiences may
converge and market segments will progressively overlap.
Keywords: sharing economy; carsharing; business-to-consumer; peer-to-peer; innovation
adoption; two-sided platform
Declarations of interest: none
Acknowledgements: Funding has been provided by Dialogic, the Rathenau Institute, and the
Dutch research council NWO under the “Sustainable Business Models” program (no. 438-14-
904). We thank TNS Nipo for providing us with the survey data.
4
1. Introduction
Next to other innovations like electric vehicles and automated vehicles, carsharing is regarded as
an important part of the transition towards a sustainable mobility system (Sperling, 2018). The
increasing diffusion of carsharing in recent years is supported by technological innovations like
smart keys, GPS services, as well as the widespread use of smartphones. However, carsharing is
still a niche phenomenon. Despite the fact that carsharing exists already for more than twenty
years in many Western countries, its current users can still be considered early adopters. In The
Netherlands, for example, roughly one percent of adults participate in some form of carsharing
(Jorritsma et al., 2015). The challenge will be to attract mainstream consumers to adopt the
carsharing service and change their mobility behavior. While the economic gains of access
instead of ownership of an expensive, non-continuously used asset like a car is clear for people
with limited car dependence, car ownership is nevertheless deeply entrenched in our societies.
Insight into the characteristics and motives of adopters and non-adopters of carsharing is thus
important for policy makers and businesses alike.
Extant research on carsharing users is quite extensive, but surveys are usually based on small
sample sizes covering only a single city. The large majority of studies is also limited in its
exclusive focus on Business-to-Consumer (B2C) carsharing. In the B2C model, otherwise
referred to as ‘classic carsharing’, an organization owns a fleet of cars which are rented out to
users. Yet, recently the Peer-to-Peer (P2P) model has gained traction, in which consumers rent
out their own cars to other consumers through a two-sided platform operated by a coordinating
carsharing organization. Our main contribution is to investigate whether characteristics and
motives differ between B2C and P2P carsharing.
5
We see a large growth in P2P carsharing as well as a blurring of business models with new
hybrid forms, offering both privately-owned and business-owned cars. This raises the question of
whether the adopters of the different types of carsharing differ in their characteristics and
motivations. The recent diffusion of P2P carsharing in which car owners act as suppliers of
shared cars also calls for more research into the characteristics and motives of P2P car providers
and how they compare to people sharing their cars informally (i.e. outside an online platform) to
friends and family. Our study thus contributes to the study of two-sided platforms in the mobility
sphere where users play a role as suppliers as well as consumers.
We analyze a unique dataset stemming from a survey of 1,836 Dutch citizens with some having
adopted carsharing, some having a self-stated interest to adopt carsharing in the next year, and
some showing no interest in adopting carsharing. The database includes socio-demographics,
modes of transport and car ownership data, as well as variables on motivations and barriers for
carsharing and attitudes towards it. The database covers the adopters of both B2C and P2P
carsharing and covers the whole of The Netherlands. We will analyze characteristics and motives
of adopters, those that declare an intent to adopt, and non-adopters. We further look into
frequency of use and car purchase avoidance among carsharing users. Regarding P2P car sharing
specifically, we also investigate car owners who supply their car on a P2P sharing platform.
We will proceed as follows. The next section provides an extensive literature review on
carsharing adopter characteristics and motives (Section 2). This is followed by the method
section (Section 3) and the result section (Section 4). The paper ends with a discussion of the
findings and some concluding remarks (Section 5).
6
2. Literature Review
2.1 Carsharing
Carsharing is a system that allows people to use locally available cars at any time and for any
duration (Frenken, 2015). Early experiments were set up in different European countries, e.g. in
Switzerland (1948, ‘Sefage’), France (1972, ‘Procotip’) and the Netherlands (1973, ‘Witkar’) but
failed to operate successfully and were suspended (Shaheen and Cohen, 2007). Successful
carsharing initiatives started in the late 1980s in Switzerland and Germany and were at first
organized in small projects of environmentally-minded groups (Shaheen et al., 1998). In Europe
carsharing has seen a continuous, yet slow growth and geographical diffusion in the 1990s and
early 2000s, but only more recently have new forms of carsharing brought a significant rise in
carsharing numbers. The first entrants in the carsharing market organized carsharing in a
Business-to-Consumer (B2C) form, in which the organization (be it for-profit or not-for-profit)
owns a fleet of cars that the customers can use. They were initially all based on a roundtrip (RT)
mode in which cars have to be returned at the end of the trip to the same spot they were rented
from. In 2008 a new form of B2C carsharing, free-floating (FF) carsharing, was introduced that
allowed the car to be dropped off anywhere in a designated operation area (Daimler, 2008). Users
can locate the car closest to them using smartphone technology (Ehrenhard et al., 2017). Around
2010, another form of organized carsharing appeared: Peer-to-Peer (P2P) carsharing. Here, the
carsharing organization provides a platform where private car owners and users can be matched
and additional services like insurances are offered (Shaheen et al., 2012). P2P carsharing can
therefore be characterized as a two-sided platform with private individuals acting both as
suppliers and consumers. Some people take on both roles, while others act as either suppliers or
consumers. Scaling the B2C carsharing business model geographically to areas with lower
7
population density is challenging. P2P carsharing can thus be an alternative in such locations, as
private car owners can make their car available in any region of a country (Münzel et al., 2017).
Carsharing in the Netherlands started with the pioneering ‘Witkar’ (‘white car’) experiment in
1973 in Amsterdam. The experiment failed to be successful and was ultimately stopped in 1988.
In the 1990s carsharing slowly grew and was stimulated by governmental organizations for the
environmental advantages it promised (Nijland and van Meerkerk, 2017). Since then expectations
around carsharing have been high, but even today it has to be seen as a niche market. In 2015 a
Green Deal between governmental authorities, companies and environmental organizations was
set up with the aim to stimulate carsharing and reach 100,000 carsharing cars by 2018
(Rijksoverheid, 2015). In early 2017 there were approximately 30,700 shared cars in the
Netherlands, with a growth of about 23 percent compared to a year before. The increase is mainly
happening in the four largest cities (Amsterdam, Rotterdam, The Hague, and Utrecht). The fastest
growth is taking place on P2P platforms, which supply 86 percent of shared cars (KpVV CROW,
2017).
2.2 Characteristics of carsharing adopters
The first consumers of an innovative service are of particular interest, since they can be decisive
for the successful diffusion of such innovation. The widely applied study by Rogers (2003)
characterizes ‘early adopters’ as “typically younger in age, having a higher social status, more
financial lucidity, advanced education, and being more socially forward than late adopters”.
Rogers’ (2003) categorization of adopters aids the understanding of human behavior in relation to
innovativeness and can thus help identify differences in characteristics as well as motivations of
carsharing adopters and provide insights about potential adopter groups.
8
Extant studies can be divided in early studies focusing on stated preferences regarding people’s
intention to adopt carsharing in the future, and a more recent wave of studies on true carsharing
adopters. In our literature review, we limit ourselves to revealed preference studies or studies that
combine stated and revealed preference as these provide more reliable insight in adoption
compared to stated preferences. Our own study focuses on true adopters as well, although we also
look into respondents who did not adopt but intend to do so in the coming year.
Based on a review of research on carsharing adoption (Table 1), it can be concluded that most
existing research on carsharing users found that socio-demographic characteristics of adopters are
consistent with the usual character traits of early adopters (e.g. Burkhardt and Millard-Ball 2006;
Wappelhorst et al. 2013; Kawgan-Kagan 2015). Drivers or barriers for adoption are less often
explored and differ mostly depending on the time of the study. Most studies do not focus on
identifying and describing carsharing adopters but rather report on socio-demographics as a side
note. Furthermore, socio-demographic characteristics as well as motives are often only described,
without analyzing the influence they have on adoption. It also needs to be noted that many of
these studies reporting on user attributes have small sample sizes, surveyed specific services or
specific locations, and usually focus on one type of carsharing. Table 1 summarizes the
descriptive findings of earlier studies on the characteristics and motives of carsharing adopters.
Age
In line with theory on early adopters of innovations, the age of carsharers is most often described
as younger than average (Becker et al., 2017; Brook, 2004; Firnkorn and Müller, 2015; Jae-Hun,
2017; Kopp et al., 2013; Lane, 2005; Martin et al., 2010; Millard-Ball et al., 2005; Mueller et al.,
2015; Prettenthaler and Steininger, 1999; SmartAgent, 2011; Steininger et al., 1996). However,
9
some studies report that people who use carsharing are especially from middle-aged groups
(Clavel and Floriet, 2009; Katzev, 2003; Knie et al., 2016; Koch, 2002; Le Vine and Polak, 2017;
Wappelhorst et al., 2013).
Gender
Results on gender differences regarding carsharing adoption are comparable to findings on the
early adopters of other innovations: most studies and reports find more males than females
participating in carsharing (bcs, 2016; Becker et al., 2017; Clavel and Floriet, 2009; Firnkorn and
Müller, 2015; Jae-Hun, 2017; Knie et al., 2016; Koch, 2002; Kopp et al., 2013; Le Vine and
Polak, 2017; Mueller et al., 2015; Wappelhorst et al., 2013), while fewer studies observe the
opposite or an equal division in gender (Brook, 2004; Katzev, 2003; Martin et al., 2010; Millard-
Ball et al., 2005).
Level of education and income
Early adopters of innovations are typically described as highly educated individuals with an
above-average income. Carsharers also typically have a high level of education (6t 2014; Becker,
Ciari, and Axhausen 2017; Brook 2004; Clavel and Floriet 2009; Clewlow 2016; Firnkorn and
Müller 2015; Katzev 2003; Kawgan-Kagan 2015; Knie et al. 2016; Koch 2002; Kopp, Gerike,
and Axhausen 2013; Lane 2005; Martin, Shaheen, and Lidicker 2010; Millard-Ball et al. 2005;
Mueller, Schmoeller, and Giesel 2015; Prettenthaler and Steininger 1999; Shaheen et al. 2004;
Steininger, Vogl, and Zettl 1996; Le Vine and Polak 2017; Wappelhorst et al. 2013) and higher
than average income (Clewlow, 2016; Firnkorn and Müller, 2015; Katzev, 2003; Kawgan-Kagan,
2015; Kopp et al., 2013; Le Vine and Polak, 2017; Millard-Ball et al., 2005; Mueller et al., 2015;
Prettenthaler and Steininger, 1999; Shaheen et al., 2004; Wappelhorst et al., 2013). Only few
10
studies found carsharers to have a medium income (Becker et al., 2017; Brook, 2004; Koch,
2002).
Household characteristics
There are different findings concerning the household size of the typical carsharer. While it is
argued that especially single or small households profit from the cost-efficient opportunity to use
a shared car, also families in larger households can choose for carsharing as a way to ease the
tight budgeting. In some studies small households are clearly identified as being overrepresented
in the group of carsharers (Knie et al., 2016; Koch, 2002; Kopp et al., 2013), whereas other
studies actually identify families to be the typical carsharers (Braun et al., 2016; Coll et al., 2014;
Le Vine and Polak, 2017). Nevertheless, there are studies that do not confirm any of the
hypotheses concerning the typical household size of carsharing adopters and do not see one group
dominating (bcs, 2016; Becker et al., 2017; Prettenthaler and Steininger, 1999). Understandably,
private car ownership is typically lower in the households of carsharing adopters (bcs, 2016;
Becker et al., 2017; Clewlow, 2016; Katzev, 2003; Knie et al., 2016; Millard-Ball et al., 2005;
Prettenthaler and Steininger, 1999; Steininger et al., 1996). Carsharing is often described as
diffusing best in high population density areas (Braun et al., 2016; Hampshire and Gaites, 2011;
Koch, 2002; Kortum, 2014; Millard-Ball et al., 2005).
Motives for carsharing
Findings on the motives of consumers to start carsharing suggest that, often, the main reasons to
join carsharing are cost-savings and increased convenience: time saving compared to using public
transport, an additional transportation mode, gaining freedom, the infrequent need of a car or a
large car in case of transporting large items (6t, 2014; Bardhi and Eckhardt, 2012; Jae-Hun,
11
2017; Katzev, 2003; Knie et al., 2016; Lane, 2005; Meijkamp, 2000; Millard-Ball et al., 2005;
Prettenthaler and Steininger, 1999; Steer Davies Gleave, 2017a, 2016, 2017b). Even though
environmentalism and sustainability are often marketed as the advantage of carsharing, these are
most often not the primary reason to join a carsharing organization (Bardhi and Eckhardt, 2012;
Lane, 2005; Millard-Ball et al., 2005). However, earlier studies did find environmental aspects to
be primary motives. For example, the study by Steininger et al. (1996) finds that environmental
aspects are high on the motivation list of carsharing members. Specific motives range from own
contribution to traffic mitigation and lower car use due to environmental concerns, a desire to
drive newer cars which are less polluting, and seeing less cars being produced. Also Koch (2002)
finds almost two-thirds of his respondents to be stating ecological motives. This is comparable to
the findings of Truffer (2003) who characterizes early carsharing adopters as ecologically
motivated and valuing the social connection between members, but also describing these factors
to be losing ground to financial and convenience motivations when the service grew.
Furthermore, multiple studies find that joining carsharing is often motivated by changes in
personal circumstances (e.g. moving house, a new job, starting a family), which may trigger a
change in people’s habits (Clavel and Floriet 2009; Kent, Dowling, and Maalsen 2017; Kent and
Dowling 2013).
P2P carsharing adoption
Despite the abundance of information about B2C carsharing users and their characteristics, little
is known about the adopters (users and providers) of two-sided P2P carsharing platforms. P2P
carsharing has only recently been introduced on the market but has grown tremendously. The
offered service differs for the users of the shared cars as compared to B2C models on e.g.
communication channels and car access, but more importantly a second adopter side is added –
12
the car provider. This group of car owners willing to share their car might differ largely from the
car user adopters and it is likely to have different motives to adopt carsharing. Wilhelms et al.
(2017a, b), Dill et al. (2017) and Shaheen et al. (2018) have studied the providers and users of
P2P carsharing in Germany, Portland (US) and across the US, respectively. All studies draw a
similar picture of the characteristics of P2P carsharing adopters compared to studies on B2C
carsharing. Dill et al. (2017) and Shaheen et al. (2018) find adopters to be younger than the
population average and Wilhelms et al. (2017a, b) to be younger or of medium age. A high level
of education is identified by all studies for both suppliers and renters. A slightly higher income is
identified for both suppliers and users (Shaheen, Martin, and Bansal 2018; Wilhelms, Henkel,
and Merfeld 2017; Wilhelms, Merfeld, and Henkel 2017), while Dill et al. (2017) finds renters to
have a lower income than the cities’ average. While Dill et al. (2017) identify more women to be
renters, Shaheen et al. (2018) find more male adopters. Wilhelms et al. (2017a, b) and Dill et al.
(2017) also found that providers have fewer cars and use them less than the average population,
and both groups to be more positive towards active transport modes (walking, biking) and public
transport. The motives of renters are partly similar to studies on B2C carsharing – all three
analyses find renters to be motivated by saving money and time as well as convenience and
mobility gains and flexibility. The possibility to live without a car and to get a specific mobility
experience or signal status are also mentioned, as well as interest in the concept of sharing and
supporting the local community. Providers were found to be motivated by cost savings and the
possibility to increase their disposable income, but also by the joy to provide a mobility option to
others, facilitating experiences by providing special cars, and an interest in sustainability. Dill et
al. (2017) report that providers had mostly positive experiences, but some of them also stated that
demand was not high enough or that the hassle was not worth it. Shaheen et al. (2018) find that
the biggest concern for providers is the fear of vehicles damage, while renters are hindered in the
13
usage by availability issues and the distance to the vehicles. Other research only explored
perceptions of the general population towards P2P. For example, Ballús-Armet et al. (2014)
surveyed people in California and found that respondents considered convenience and availability
as the main attractions of becoming renters, as well as economic benefits. A quarter of
respondents considered renting out their vehicle but were concerned of liability issues as well as
convenience and availability.
14
Table 1: Descriptive findings on characteristics of carsharing adopters
Age Gender Education Income Household Car ownership Location Motives B2C Users
Young (Becker et al., 2017; Brook, 2004; Firnkorn and Müller, 2015; Jae-Hun, 2017; Kopp et al., 2013; Lane, 2005; Martin et al., 2010; Millard-Ball et al., 2005; Mueller et al., 2015; Prettenthaler and Steininger, 1999; SmartAgent, 2011; Steininger et al., 1996) Middle aged (Clavel and Floriet, 2009; Katzev, 2003; Knie et al., 2016; Koch, 2002; Le Vine and Polak, 2017; Wappelhorst et al., 2013)
More male (bcs, 2016; Becker et al., 2017; Clavel and Floriet, 2009; Firnkorn and Müller, 2015; Jae-Hun, 2017; Knie et al., 2016; Koch, 2002; Kopp et al., 2013; Le Vine and Polak, 2017; Mueller et al., 2015; Wappelhorst et al., 2013) More female (Martin et al., 2010; Millard-Ball et al., 2005) Equal division (Brook, 2004; Katzev, 2003)
Highly educated (6t, 2014; Becker et al., 2017; Brook, 2004; Clavel and Floriet, 2009; Clewlow, 2016; Firnkorn and Müller, 2015; Katzev, 2003; Kawgan-Kagan, 2015; Knie et al., 2016; Koch, 2002; Kopp et al., 2013; Lane, 2005; Le Vine and Polak, 2017; Martin et al., 2010; Millard-Ball et al., 2005; Mueller et al., 2015; Prettenthaler and Steininger, 1999; Shaheen et al., 2004; Steininger et al., 1996; Wappelhorst et al., 2013)
High income (Clewlow, 2016; Firnkorn and Müller, 2015; Katzev, 2003; Kawgan-Kagan, 2015; Kopp et al., 2013; Le Vine and Polak, 2017; Millard-Ball et al., 2005; Mueller et al., 2015; Prettenthaler and Steininger, 1999; Shaheen et al., 2004; Wappelhorst et al., 2013) Medium income (Becker et al., 2017; Brook, 2004; Koch, 2002)
Small households (Knie et al., 2016; Koch, 2002; Kopp et al., 2013) Families (Braun et al., 2016; Coll et al., 2014; Le Vine and Polak, 2017) All sizes (bcs, 2016; Becker et al., 2017; Prettenthaler and Steininger, 1999)
Lower car ownership (bcs, 2016; Becker et al., 2017; Clewlow, 2016; Katzev, 2003; Knie et al., 2016; Millard-Ball et al., 2005; Prettenthaler and Steininger, 1999; Steininger et al., 1996)
Higher population density (Braun et al., 2016; Hampshire and Gaites, 2011; Koch, 2002; Kortum, 2014; Millard-Ball et al., 2005)
Cost-savings and convenience-gaining (6t, 2014; Bardhi and Eckhardt, 2012; Jae-Hun, 2017; Katzev, 2003; Knie et al., 2016; Lane, 2005; Meijkamp, 2000; Millard-Ball et al., 2005; Prettenthaler and Steininger, 1999; Steer Davies Gleave, 2017a, 2016, 2017b) Environmental motives (Koch, 2002; Steininger et al., 1996; Truffer, 2003) Changes in personal circumstances (Clavel and Floriet, 2009; Kent et al., 2017; Kent and Dowling, 2013)
P2P Users Young (Dill et al., 2017) Medium to younger age (Wilhelms et al., 2017b, 2017a)
More female (Dill et al., 2017)
Highly educated (Dill et al., 2017)
Higher income (Wilhelms et al., 2017b, 2017a) Lower income (Dill et al., 2017)
Money and time savings, convenience, gaining mobility (Ballús-Armet et al., 2014; Dill et al., 2017; Shaheen et al., 2018; Wilhelms et al., 2017b, 2017a)
P2P Suppliers Young (Dill et al., 2017)
Highly educated (Dill et al., 2017)
Lower car ownership (Dill et al., 2017)
Money gains, providing to others and sustainability
15
(Dill et al., 2017; Shaheen et al., 2018)
Both P2P Users and Suppliers Young (Shaheen et al., 2018)
More male (Shaheen et al., 2018)
Highly educated (Shaheen et al., 2018)
Slightly higher income (Shaheen et al., 2018)
16
2.3 Factors influencing carsharing adoption
In addition to descriptive characterizations of carsharing adopters and their motives, it is of great
importance to identify the variables that actually influence carsharing adoption using multivariate
analysis. The findings presented in Section 2.2 only provide descriptive statistics about
carsharing adopters without analyzing which variables actually have an effect on the likelihood of
being an adopter in a multivariate setting, i.e. while controlling for other factors. Table 2 provides
an overview of multivariate studies that have analyzed in different settings which socio-
demographic variables influence carsharing adoption (or membership duration, or use frequency).
Again, we exclude studies that are based solely on intentions.
Almost all studies find age to significantly influence the likelihood of someone adopting
carsharing, with most studies stating that younger age positively or older age negatively
influences adoption (Becker et al., 2017; Dias et al., 2017; Habib et al., 2012; Hahn, 2015;
Meijkamp, 2000). Gender is identified as an influencing variable in some studies as well, with
more studies stating being male to have a positive influence on adoption (Becker et al., 2017;
Hahn, 2015; Juschten et al., 2017) than being female (Kim et al., 2015). Habib et al. (2012) find
that membership duration is positively influenced by being female, while use frequency is
influenced by being male. Becker et al. (2017) identify the positive effect of being male for
freefloating carsharing, but not for the roundtrip model.
All studies that found a significant effect of education on carsharing adoption agree that a higher
level of education positively influences adoption (Becker et al., 2017; Dias et al., 2017; Hahn,
2015; Juschten et al., 2017). Results on income, on the other hand, go in both directions: a higher
17
income having a positive influence is discovered by multiple studies (Becker et al., 2017; Dias et
al., 2017; Hahn, 2015; Juschten et al., 2017), while Kim et al. (2015) identify the opposite.
Household size is identified to be of significant influence in a few studies, but with mixed results.
Most studies state that living alone or in small households positively influences adoption (Habib
et al., 2012; Kim et al., 2015), whereas Dias et al. (2017) report the opposite. Agreement is found
on the influence of car ownership and the number of cars in a household: living without a car or
owning fewer cars positively influences adoption or the intention to do so (Becker et al., 2017;
Dias et al., 2017; Habib et al., 2012; Hahn, 2015; Juschten et al., 2017).
A few studies include attitudes in the analysis and identify environmental concern, a
dissatisfaction with transport availability, an orientation towards public transport and perceptions
on financial saving possibilities to have a positive influence on adoption (Becker, Ciari, and
Axhausen 2017; Meijkamp 2000; Shaheen 1999).
The review of existing literature on the profile and motives of carsharing adopters reveals that
most research has focused on only one business model and provider (B2C mostly) and/or one
geographical area (often one city). A comparison of different carsharing types and the analysis of
a large geographical area including urban and more rural regions is missing. A notable exception
is the study by Prieto et al. (2017), which examined adoption intentions of B2C and P2P
carsharing, yet it does not provide insights on actual adopters. They find that P2P carsharing
appears to be favorable to a larger range of consumers and argue that P2P carsharing offers an
easier and more flexible service and can thus be more attractive than B2C carsharing.
18
Table 2: Multivariate analyses on carsharing adoption
Study Location, Data source
Dependent Variable
Age Gender Education Income, Work
Household size
Transport in household
Location Attitudes
(Habib et al., 2012)
Montreal, Canada. B2C users
Membership duration
Younger +
Female +
Larger - Car ownership -
Use frequency Middle aged +
Male +
(Shaheen, 1999)
California, USA. B2C users
Adoption interest CS
High environmental concern +, Dissatisfaction with transport availability +
(Hahn, 2015)
Germany. General transportation survey including B2C users
CS membership
Younger +
Male + Higher + Higher Income +
Car ownership -
(Becker et al., 2017)
Switzerland. B2C roundtrip and freefloating users and driver license holders
CS adoption Younger + for ff
Male + for ff
Higher + Carless hh + Bad transit access + for ff
Transit orientation +
Frequency of use
Younger + for ff
Male + for ff
Higher Income + for ff
Carless hh + Transit card -
(Meijkamp, 2000)
Netherlands. B2C users
Attitude towards CS
Older - Saving perceptions +
(Dias et al., 2017)
Puget Sound, WA, USA. Travel Survey including B2C users
CS adoption Older - Higher + Low Income - Full time employment +
Single hh -Number of children -
Number of cars -
(Kim et al., 2015)
Seoul, South Korea. B2C users
Attitude toward continuing CS use
Older + Female +
Higher Income - Students+ Non-office workers +
Single hh +
(Juschten et al., 2017)
Switzerland. National Travel Survey + B2C booking data.
CS membership
Middle aged +
Male + Higher + Higher income +
Lower car/license holder ratio +
(Prieto et al., 2017)*
London, Madrid, Paris, Tokyo. Survey of car owners.
Adoption intention any CS type
Older - Male + Higher + Being main driver in hh -
City center +
Adoption intention B2C
Older - Male + Higher + Fulltime - Single - Being main driver in hh -
City center +
Adoption Male + Single + City center
19
intention P2P +
+ positive influence; - negative influence; CS=carsharing; ff=free-floating carsharing; hh=household. *=despite being based on intentions only we included Prieto et al. because it analyses P2P-B2C differences.
20
3. Methodology
In our study, we make use of a survey dataset collected by the Dutch market research agency
TNS Nipo in 2014. TNS Nipo accessed a panel of about 140,000 people representative of the
Dutch population, and used a pre-screening to all people over the age of 18 to be able to over-
sample on true car sharing adopters. The survey was answered by 1,835 participants of which 258
respondents participate in at least one form of organized carsharing, 221 participants stated to
plan on starting carsharing in the next 12 months, while the rest showed no interest in doing so.
The survey includes 151 B2C carsharing users, 87 P2P carsharing users and another 46 that use
both. In the analysis, the latter cases are duplicated leading to a total of 197 respondents using
B2C carsharing and 133 using P2P carsharing.
When regarding the supply side, the database includes 42 cases of supplying cars on a P2P
carsharing platform; 498 participants stated that they do lend out their car but in a private manner
to family and friends, while 934 participants own a car but do not supply it. Socio-demographics,
attitudes and motivations of the respondents are analyzed and descriptive results are reported that
characterize the adopters of the different carsharing forms, as well as potential adopters and non-
adopters. Furthermore, multinomial logistic regressions (MLR) analyses are performed to
identify which variables influence the chance to be part of one group over another group.
Multiple MLRs are used to identify influencing variables for the groups of adopters, potential
adopters and non-adopters as well as frequent versus infrequent users; adopters avoiding a car
purchase through carsharing or not; P2P versus B2C users; and P2P suppliers, private suppliers
and non-suppliers (see Table 3). The data is checked for multicollinearity and no issues were
identified. Correlation tables can be found in the appendix. In addition to MLR, ANOVA
analyses were conducted to compare the groups and the results reveal similar relationships.
21
Table 3: The dependent variables and case numbers (total number of respondents = 1,835)
MLR Analyses of groups Categories n
1 Adoption
Adopter of B2C and/or P2P carsharing (be it as user or supplier)
258
Potential adopter (interested to begin carsharing in next 12 month)
221
Non-adopter (not interested to begin carsharing in next 12 months)
1,356
2 Frequency of use Frequent carsharing users (more than five times per year) 113 Infrequent carsharing users (five times per year or less) 125
3 Car purchase decision Would have bought a car if had not started carsharing 34 Would have maybe bought a car 70 Would not have bought a car 76
4 Carsharing type
User of B2C carsharing 151 User of P2P carsharing 87 User of both forms (cases are duplicated and coded as B2C and P2P users in the MLR analysis)
46
5 Car suppliers Car supplier on P2P platform 42 Private car supplier (to friends/family) 498 Car owner but not supplying 934
The independent variables
Based on the earlier findings on adopters of carsharing presented in the literature review (Section
2) and considering the data availability of the survey database, variables were chosen to be
included in the MLR analyses. Based on Table 1 and 2, several socio-demographic variables are
included in all MLR analyses: age, gender, income category, education level, and a dummy
variable if children live in the household1. Furthermore, a dummy variable is included covering
whether the respondent lives in one of the four largest cities of the Netherlands (Amsterdam,
Rotterdam, The Hague, Utrecht) indicated as G4 cities. The accessibility of transportation options
is characterized through including a dummy variable on the ownership of a public transport
subscription and a dummy variable if there is no car available in the household. Environmental
1 Variable is highly correlated with the variable household size.
22
attitudes are analyzed through including a dummy variable if the respondent voted for a green
party in the last general election.
The first MLR comparing the adopter and non-adopter groups includes all these variables. The
second, third and fourth MLR which analyze only current carsharing adopters additionally
include variables on the motivations that led respondents to adopt carsharing. Three dummy
variables are included that represent cost, convenience or the environment as the most important
reasons to carshare. In the analyses on the frequency of use and the car purchase decision a
dummy variable on the type of carsharing (B2C or P2P) is included to control for the influence of
the type of carsharing on the frequency and car purchase decision respectively. In the fifth MLR
that compares car owners and their adoption of P2P carsharing as suppliers, motivations cannot
be included since the analysis includes groups of non-adopters. The variable on car ownership is
redundant and excluded. In order to control for the perceived trust or safety in a municipality
influencing the decision to lend a car to others, an additional variable on the number of crimes
per 1,000 inhabitants is included in the analysis on car supplier groups. Table 4 illustrates which
variables are included in which MLR analysis.
Table 4: Independent variables
Users Suppliers MLR # 1
Adoption 2 Frequency of use
3 Car purchase decision
4 Carsharing type
5 Car suppliers
Socio-demographics
Age X X X X X Gender X X X X X Income level (27 levels)
X X X X X
Education level (8 levels)
X X X X X
Living in G4 city X X X X X Children in household
X X X X X
Transportation No car in X X X X
23
set household Public transport subscription
X X X X X
Attitude Green party voter X X X X X Control Trust Crime numbers X
Motivation
Most important reason: Cost
X X X
Most important reason: Convenience
X X X
Most important reason: Environment
X X X
Control CS type
P2P or B2C user X X
4. Results
4.1 Descriptive characterization of carsharers The Dutch carsharing users of our sample are similar in their socio-demographics to earlier
studies and partly to early adopters of innovations in general. Table 5 presents the descriptive
statistics. We observe an equal amount of males and females adopting carsharing and a mean age
of 46 years, ranging from 19 to 85 and spreading quite equally between the age groups of 25 to
70. Household income and level of education are high. Carsharing users often live in densely
populated areas, many in the four largest city of the Netherlands, and are part of households with
children. Half of the carsharing users live in car-free households and two-thirds have a public
transport subscription; amounts that are significantly higher than those of the other two groups
(i.e. potential carsharing adopters and non-adopters). Stronger attitudes towards the environment
can be observed in carsharing users: 18% of carsharing users have voted for a green party in the
last general election. At the same time, the environment is not the main motivation for carsharing
according to the majority of respondents. Only 9% of respondents mention it as the most
important reason to adopt carsharing, while 40% stated that the most important reason to carshare
24
are the cost savings and 11% stated the convenience of not owning a car. The reasons why
potential adopters have not adopted carsharing so far are manifold: while many mention that they
still have their own car or that it just has not happened yet (40% and 24% of respondents
respectively), others also state that they are missing information on the option or can easily get a
car from family or friends. Potential adopters state that they would start with carsharing when car
ownership becomes too expensive (46%) or when they would need to buy a new car (25%).
Table 5: Descriptive Statistics of adopter groups
Mean age
Male Mean education level (8 levels)
Mean income level (27 levels)
Children in household
Green party voters
Public transport subscription holders
Carfree households
Living in G4 city
Carsharing adopter
45.6 51% 6.5 14.3 29% 18% 65% 50% 42%
Potential adopter
47.0 41% 6.0 13.4 22% 9% 47% 24% 33%
Non-adopter
51.0 46% 5.7 14.0 23% 6% 37% 14% 24%
The car owners that rent out their car through a P2P platform are more often male (60%) and
have a mean age of 45. Level of education and income are again high and a large number of P2P
suppliers live with children in the household. 14% of car suppliers have voted for a green party.
These numbers are higher than for people lending out their car only privately or not at all (Table
6). The main points of concern of people not providing their car on a P2P platform are the risk of
damage (28%), that they need more information first (27%) or that they need their car too often
themselves (27%).
Table 6: Descriptive statistics of provider groups
Mean age
Male Mean education level (8 levels)
Mean income level (27 levels)
Children in household (Mean)
Green party voters
Public transport subscription holders
Living in G4 city
P2P provider 45.4 60% 6.6 14.9 40% 14% 50% 29%
Private provider
52.3 43% 5.9 14.1 25% 7% 37% 23%
Car owner but not providing
50.7 49% 5.7 14.7 26% 5% 35% 19%
25
4.2 Multinomial Logistic Regressions The first MLR explores which variables influence the chance to be not interested in doing
carsharing compared to the groups of carsharing adopters and potential adopters (Table 7). The
group of not interested respondents acts as the reference category of the dependent variable. The
odd ratios of the carsharers and the potential adopter group have to be thus each interpreted
compared to this group. The Nagelkerke indicator (a Pseudo R²) lies at 0.21 which means that
21% of the variance is explained through the included variables. Apart from the variable on
living in the G4 cities all variables are significant when comparing carsharers and the not-
interested group. When comparing age between the non-adopters and adopters of carsharing a
significant influence is observed: age increasing by one decreases the likelihood of a respondent
being a carsharer by 1%. A younger person is thus more likely to be a carsharer. For males the
chances increase by a factor of 2.04 for being a carsharer rather than being not interested
compared to females. The education level increasing by one level increases the likelihood to be a
carsharer by 27%, the income level increases it by 6%. Having children in a household leads to a
77% increase in likelihood to be adopting carsharing. Similar to other studies (Shaheen 1999;
Efthymiou et al. 2013; Becker et al. 2017), we find a positive influence of environmental
mindedness and the use of public transportation on carsharing adoption. Voting for a green party
more than doubles the chance of being a carsharer rather than being not interested. Having a
public transport subscription as well as living in a carfree household also increases the likelihood
of being a carsharer by a factor of 2.52 and 5.46 respectively. These results show that carsharing
adopters are a clearly distinct group of people. The variables that significantly influence the
likelihood for a person to be an adopter are in line with Rogers’ early adopter description as well
as with findings from most previous studies. However, in contrast to findings that carsharing
diffuses best in dense city areas (e.g. Millard-Ball et al. 2005; Hampshire and Gaites 2011) we do
26
not find an influence of living in one of the G4 cities. With our large dataset, including
respondents from more rural areas, we can thus not support the statement that carsharing is only a
city phenomenon. When comparing the not-interested group to the potential adopters, only age
has a significant influence on the likelihood of belonging to one group. For an increase of one
year in age the likelihood to be a potential adopter decreases by 2%. This is consistent with the
trend that the younger urban generation is more open to gaining access to assets rather than
owning them (Rifkin, 2015).
Table 7: MLR #1 Adopters, potential adopters vs. non-adopters
Reference category: Not interested in carsharing Carsharer Potential carsharer Exp(B) Exp(B) Age 0.99** 0.98*** Gender (Male) 2.04*** 0.98 Education level 1.27*** 1.08 Household income 1.06** 0.98 Dummy Children (<18) in household 1.77*** 0.85 Dummy Green party voter 2.18*** 1.11 Dummy Public transport subscription 2.52*** 1.30 Dummy Carfree household 5.46*** 1.11 Dummy G4 cities 1.17 1.27
N 1292 Nagelkerke 0.21
*significance at the 0.1 level; ** significance at the 0.05 level; ***significance at the 0.01 level
Carsharing is not the main means of transport for most carsharers but rather another option in a
wider set of transportation modes in use. Many carsharing organizations ask no monthly or
annual subscription fees and do not all require a registration fee which makes it easy for people to
sign up for the service and rarely use it. Therefore, it is valuable to analyze which groups of
people use carsharing as a regular part of their transportation mix and who uses carsharing only
as an infrequent additional mode of transport for special circumstances. The second MLR thus
explores which variables influence the frequency of use (Table 8). Infrequent users are
27
characterized as adopters using carsharing five times per year or less, while frequent users are
respondents stating using carsharing more than five times per year. The studies by Becker et al.
(2017) and Habib et al. (2012) showed that frequency of use is higher for younger to middle
aged, male adopters who have higher incomes and live in a carless household. In contrast, our
analysis only finds that being a B2C user increases the likelihood to be a high frequent user by a
factor of 17.8 and living in a carfree household, not surprisingly, increases it by a factor of 2.7.
Frequent users thus tend to prefer the more convenient and professional B2C service over the P2P
service.
Table 8: MLR #2 Frequency of use
Reference category: Low frequency use High frequency use Exp(B) Age 1.01 Gender (Male) 1.00 Education level 1.12 Household income 1.02 Dummy Children (<18) in household 1.85 Dummy Green party voter 0.92 Dummy Public transport subscription 1.61 Dummy Carfree household 2.68** Dummy G4 cities 1.27 Most important reason to carshare: Costs 0.83 Most important reason to carshare: Convenience 2.46 Most important reason to carshare: Environment 2.54 Dummy B2C user 17.80*** Dummy P2P user 1.04
N 180 Nagelkerke 0.307
*significance at the 0.1 level; ** significance at the 0.05 level; ***significance at the 0.01 level
The third MLR analyses the factors influencing the avoidance of car purchase. Table 9 shows that
only car ownership and the motive to carshare influence whether a car purchase is prevented
through carsharing. Respondents living in a carfree household are 94% less likely to state that
they would have bought a car if they had not started carsharing, meaning that especially
28
households who own at least one car avoided a car purchase through using carsharing. Stating
costs as the first reason to adopt carsharing reduces the likelihood of preventing a car purchase by
70%, illustrating that for the cost-sensitive adopters buying a car as an alternative to carsharing is
not an option, possibly because household finances do not allow this in any case.
Table 9: MLR #3 Car purchase decision
Reference category: Would not have bought a car if had not started carsharing
Would have bought a car if had not started carsharing
Would have maybe bought a car if had not started carsharing
Exp(B) Exp(B) Age 0.98 0.98 Gender (Male) 0.94 1.30 Education level 0.86 0.79 Household income 1.13 1.08 Dummy Children (<18) in household 1.10 1.26 Dummy Green party voter 1.42 1.07 Dummy Public transport subscription 0.68 1.22 Dummy Carfree household 0.06*** 0.30** Dummy G4 cities 0.80 0.93 Most important reason to carshare: Costs 0.30* 2.23* Most important reason to carshare: Convenience 0.31 2.64 Most important reason to carshare: Environment 1.25 7.11** Dummy B2C user 0.50 1.17 Dummy P2P user 0.57 1.23
N 180 Nagelkerke 0.407
*significance at the 0.1 level; ** significance at the 0.05 level; ***significance at the 0.01 level
The fourth MLR explores which variables explain for a respondent the preference to be a B2C
user over a P2P user (Table 10). The socio-demographic differences between the two user
groups are small. The only significant influence on the chance to choose P2P over B2C
carsharing stems from the income variable and the variable of convenience being the most
important reason to adopt B2C carsharing. Increasing the income level by 1 increases the chance
for being part of the B2C user group by 9%. Seeing convenience as the most important reason to
29
carshare increases the likelihood of being a B2C user compared to a P2P user by a factor of 3.25.
The adopters of the two different carsharing types are thus very similar and the choice of one
form over the other is only explained by a higher income that makes it easier to pay for the more
expensive form of carsharing (B2C) that is perceived as offering more convenience (e.g. no key
exchange dates have to be arranged, clear communication channels). For all other variables no
significant influence is identified, which means that the same type of people are attracted to the
different services. This is in contrast to the results of Prieto et al. (2017) who find that older age,
higher education and fulltime employment is only influencing the adoption intention for B2C
carsharing but not for P2P carsharing.
Table 10: MLR #4 Carsharing type
Reference category: P2P user B2C user Exp(B) Age 0.99 Gender (Male) 1.72 Education level 1.02 Household income 1.09** Dummy Children (<18) in household 0.72 Dummy Green party voter 0.82 Dummy Public transport subscription 1.30 Dummy Carfree household 1.74 Dummy G4 cities 1.60 Most important reason to carshare: Costs 1.25 Most important reason to carshare: Convenience 3.25* Most important reason to carshare: Environment 1.16
N 212 Nagelkerke 0.131
*significance at the 0.1 level; ** significance at the 0.05 level; ***significance at the 0.01 level
The fifth MLR explores factors influencing a respondent to be a car supplier on a P2P platform,
a car owner privately lending or a car owner that is not providing a car (Table 11). When
comparing the group of car owners that do not lend out their car to car providers on a P2P
platform we observe an influence of gender, education level, green voting and having a public
30
transport subscription. For males the odds increase by a factor of 2.27 for being a P2P platform
provider rather than not providing compared to females. An increase in the education level by 1
increases the likelihood of being a P2P platform provider by 45%. Green voting increases the
likelihood of being a platform provider by a factor of 2.8, similar to having a public
transportation subscription, which increases the likelihood by a factor of 2.83. Therefore, also the
insights gained on P2P providers are in line with some of the typical early adopter characteristics
and findings of earlier studies on providers of P2P vehicles. Age, on the other hand, does not
have explaining power in the provider case, possibly because car ownership is required to
become a provider and is more common in higher age groups.When comparing private suppliers
(supplying only to friends and family) to not-providers, we see effects of age, gender, education
and income. Particularly interesting is the gender effect, suggesting that while men are more
prone to supply the car on a P2P platform, women choose more often to lend out the car to family
and friends in an informal way.
Table 11: MLR #5 Providers
Reference category: Not providing but car owner P2P platform provider Private provider Exp(B) Exp(B) Age 0.98 1.02*** Gender (Male) 2.27* 0.80* Education level 1.45** 1.17*** Household income 1.01 0.94*** Dummy Children (<18) in household 1.45 1.09 Dummy Green party voter 2.80* 1.03 Dummy Public transport subscription 2.83*** 0.87 Dummy G4 cities 1.30 1.02 Crime rate in municipality 1.00 1.00
N 1023 Nagelkerke 0.071
*significance at the 0.1 level; ** significance at the 0.05 level; ***significance at the 0.01 level
31
5. Concluding remarks
Based on the results of this study, four main findings can be highlighted: (1) Adopters of
carsharing constitute a rather distinct group of people whose characteristics are consistent with
those described in early adopter theory and existing carsharing research. (2) B2C and P2P users
differ only slightly, showing that carsharing business models are different mostly for the
suppliers and not so much for users. (3) Interesting differences in attitudes exist for adopter
groups. Environmental attitudes as well as gender are explaining differential carsharing choices.
(4) Carsharing acts as a substitution to a second car in a household. We elaborate on these
findings below:
First, the unique dataset analyzed in this study, including next to carsharing adopters, also
potential adopters and non-adopters, makes it possible to draw the conclusion that adopters are a
clearly distinct group of people. The variables that significantly influence the likelihood for a
person to be an adopter are in line with Rogers’ early adopter description as well as with findings
from most previous studies: carsharing is especially popular among younger people, higher
educated people, and higher income households. However, innovation adoption theories need to
be extended for the new role category of ‘supplier-users’ on two-sided platforms in e.g. the
sharing economy. Findings of this research suggest that the roles of users and supplier-users are
distinctively different and motives, barriers and attributes need to be identified in order to being
able to foster a balanced level of demand and supply.
Second, no large differences can, on the other hand, be identified for B2C and P2P users.
Although the B2C and P2P organizations differ greatly in their business model, the users of the
B2C and P2P carsharing services are rather similar. Users seems first and foremost interested in
getting access to a car rather than the business model supplying it. Indeed, the two services can be
32
considered close substitutes. The key difference holds that B2C users are more frequent users that
look for a convenient service, whereas P2P users have less money to spend and use carsharing
only infrequently. It can be followed that at the moment P2P carsharing is rather used for special
purposes, whereas B2C users make it part of their normal routine. Since most B2C providers in
the Netherlands charge some kind of fixed fees, B2C users make a more financially committed
choice and only become a member of the service when they know they need it. Besides income
and the convenience motive other socio-demographic aspects are the same. This finding is
important in the light of the ongoing convergence of the two business models foreshadowed by
some platforms starting to offer both services (e.g., MyWheels) or P2P platforms getting into
alliances with traditional rental companies (e.g., SnappCar with Europcar). Furthermore, P2P
carsharing is becoming more convenient and professionalized2 through smart locks that make the
key handover redundant, while at the same time B2C carsharing is becoming cheaper and better
connected nationally and internationally. This suggests that, progressively, the convenience levels
and price differences will become smaller.
Third, we also gained a number of insights into gender and environmental attitudes influencing
carsharing adoption. In line with early adopter theory, being male positively influences the
likelihood of adopting carsharing and an environmental attitude positively influences the
likelihood of adopting carsharing, which is often framed as a sustainable innovation, reducing car
ownership and driven kilometers. Adopting carsharing as a supplier on a P2P carsharing platform
is also affected by pro-environmental attitudes and male gender. Men, who are more often
responsible for car management and finances in a household, see carsharing as an economic
2 This development can be compared to the development of the accommodation platform Airbnb, which has professionalized tremendously, making it a more convenient experience for users (Benner, 2017).
33
opportunity and adopt it, while women rather stick to the traditional way of sharing, lending their
car to families and friends.
Fourth,our analysis also yields that carsharing can act as a substitution for a second car in a
household. Especially households that already own a car avoided a car purchase through adopting
carsharing. Next to that carsharing can act as a car access option for cost-sensitive households
that are not able to afford car ownership. These impacts of carsharing have not yet been clearly
indicated by any of the earlier studies.
For multiple reasons there is a lot of room for carsharing to grow, so research into the current
adoption patterns and possible future developments is highly valuable. The current adopters of
carsharing seem to follow the typical early adopter characteristics of being younger and having a
higher status which has a trendsetting power that can be utilized. Yet, the study also shows that
cost and convenience are the main motivational factors for people to adopt carsharing, which
suggests that carsharing can reach out to the mass market made up by people lacking a strong
environmental attitude. Diffusion is further supported by the convergence of business models
rendering the more convenient B2C services cheaper and the cheaper P2P service more
convenient.
This leads us to the policy implication of our study. Given the large overlap in characteristics and
motives among B2C and P2P adopters, as well as the progressive convergence and competition
among carsharing business models, government policies towards carsharing can be generic. For
example, one can think of increasing the cost of car ownership (including residential parking),
preferential driving lane access for shared cars, and generic campaigns covering all carsharing
options.
34
References
6t, 2014. One-way carsharing: which alternative to private cars? Executive Summary. Paris.
Ballús-Armet, I., Shaheen, S., Clonts, K., Weinzimmer, D., 2014. Peer-to-peer carsharing: Exploring public perception and market characteristics in the San Francisco Bay area, California. Transp. Res. Rec. 2416, 27–36. https://doi.org/10.3141/2416-04
Bardhi, F., Eckhardt, G.M., 2012. Access-Based Consumption: The Case of Car Sharing. J. Consum. Res. 39, 000–000. https://doi.org/10.1086/666376
bcs, 2016. CarSharing fact sheet Nr. 2.
Becker, H., Ciari, F., Axhausen, K.W., 2017. Comparing car-sharing schemes in Switzerland: User groups and usage patterns. Transp. Res. Part A Policy Pract. 97, 17–29. https://doi.org/10.1016/j.tra.2017.01.004
Benner, K., 2017. Airbnb Tries to Behave More Like a Hotel. New York Times.
Braun, A., Koch, A., Hochschild, V., 2016. Intraregionale Unterschiede in der Carsharing-Nachfrage. disP - Plan. Rev. 52, 72–85. https://doi.org/10.1080/02513625.2016.1171051
Brook, D., 2004. Carsharing: startup issues and new operational models. Proc. Transp. Res. Board Annu. Meet.
Burkhardt, J., Millard-Ball, A., 2006. Who Is Attracted to Carsharing? Transp. Res. Rec. J. Transp. Res. Board 1986, 98–105. https://doi.org/10.3141/1986-15
Clavel, R., Floriet, M., 2009. The Increasing Development of Carsharing in France.
Clewlow, R.R., 2016. Carsharing and sustainable travel behavior: Results from the San Francisco Bay Area. Transp. Policy 51, 158–164. https://doi.org/10.1016/j.tranpol.2016.01.013
Coll, M.-H., Vandersmissen, M.-H., Thériault, M., 2014. Modeling spatio-temporal diffusion of carsharing membership in Québec City. J. Transp. Geogr. 38, 22–37. https://doi.org/10.1016/j.jtrangeo.2014.04.017
Daimler, 2008. Daimler startet Mobilitätskonzept für die Stadt: Car2go - so einfach wie mobiles Telefonieren [WWW Document]. URL http://media.daimler.com/marsMediaSite/ko/de/9914253 (accessed 1.30.17).
Dias, F.F., Lavieri, P.S., Garikapati, V.M., Astroza, S., Pendyala, R.M., Bhat, C.R., 2017. A behavioral choice model of the use of car-sharing and ride-sourcing services. Transportation (Amst). 1–17. https://doi.org/10.1007/s11116-017-9797-8
Dill, J., Mcneil, N., Howland, S., 2017. Peer-To-Peer Carsharing: Short-Term Effects on Travel Behavior in Portland, OR, Transportation Research and Education Center. Portland, OR. https://doi.org/10.15760/trec.172
35
Ehrenhard, M., Wijnhoven, F., van den Broek, T., Zinck Stagno, M., 2017. Unlocking how start-ups create business value with mobile applications: Development of an App-enabled Business Innovation Cycle. Technol. Forecast. Soc. Change 115, 26–36. https://doi.org/10.1016/j.techfore.2016.09.011
Firnkorn, J., Müller, M., 2015. Free-floating electric carsharing-fleets in smart cities: The dawning of a post-private car era in urban environments? Environ. Sci. Policy 45, 30–40. https://doi.org/10.1016/j.envsci.2014.09.005
Frenken, K., 2015. Towards a prospective transition framework. A co-evolutionary model of socio- technical transitions and an application to car sharing in The Netherlands, in: International Workshop on the Sharing Economy. Utrecht.
Habib, K.M.N., Morency, C., Islam, M.T., Grasset, V., 2012. Modelling users’ behaviour of a carsharing program: Application of a joint hazard and zero inflated dynamic ordered probability model. Transp. Res. Part A Policy Pract. 46, 241–254. https://doi.org/10.1016/j.tra.2011.09.019
Hahn, C., 2015. Eine CarSharing-Zielgruppenanalyse der großen SrV-Vergleichsstädte. TU Dresden.
Hampshire, R., Gaites, C., 2011. Peer-to-Peer Carsharing: Market analysis and potential growth. Transp. Res. Rec. J. Transp. Res. Board 2217, 119–126. https://doi.org/10.3141/2217-15
Jae-Hun, J., 2017. Motives for Participating in Sharing Economy: Intentions to Use Car Sharing Services. J. Distrib. Sci. 15, 21–26. https://doi.org/10.15722/jds.15.2.201702.21
Jorritsma, P., Harms, L., Berveling, J., 2015. Mijn auto, jouw auto, onze auto Deelautogebruik in Nederland: omvang, motieven en effecten. Den Haag.
Juschten, M., Ohnmacht, T., Thao, V.T., Gerike, R., Hössinger, R., 2017. Carsharing in Switzerland: identifying new markets by predicting membership based on data on supply and demand. Transportation (Amst). 1–24. https://doi.org/10.1007/s11116-017-9818-7
Katzev, R., 2003. Car Sharing: A New Approach to Urban Transportation Problems. Anal. Soc. Issues Public Policy 3, 65–86. https://doi.org/10.1111/j.1530-2415.2003.00015.x
Kawgan-Kagan, I., 2015. Early adopters of carsharing with and without BEVs with respect to gender preferences. Eur. Transp. Res. Rev. 7. https://doi.org/10.1007/s12544-015-0183-3
Kent, J., Dowling, R., Maalsen, S., 2017. Catalysts for transport transitions: Bridging the gap between disruptions and change. J. Transp. Geogr. 60, 200–207. https://doi.org/10.1016/j.jtrangeo.2017.03.013
Kent, J.L., Dowling, R., 2013. Puncturing automobility? Carsharing practices. J. Transp. Geogr. 32, 86–92. https://doi.org/10.1016/j.jtrangeo.2013.08.014
Kim, D., Ko, J., Park, Y., 2015. Factors affecting electric vehicle sharing program participantsâ€TM attitudes about car ownership and program participation. Transp. Res. Part D 36, 96–106. https://doi.org/10.1016/j.trd.2015.02.009
36
Knie, A., Lenz, B., Nobis, C., Nowack, F., Giesel, F., Blümel, H., 2016. Wirkung von E-Car Sharing Systemen auf Mobilität und Umwelt in urbanen Räumen (WiMobil) - Gemeinsamer Abschlussbericht.
Koch, H., 2002. User Need Report. Moses (Mobility Services for Urban Sustainability) Project. Bremen.
Kopp, J., Gerike, R., Axhausen, K.W., 2013. Status Quo and Perspectives for CarSharing Systems: the Example of DriveNow, in: Hülsmann, F., Gerike, R., Roller, K. (Eds.), Strategies for Sustainable Mobilities - Opportunities and Challenges. Ashgate, Burlington, pp. 207–226.
Kortum, K., 2014. Driving smart: Carsharing mode splits and trip frequencies., in: Transportation Research Board 93rd Annual Meeting. p. No. 14-4009.
KpVV CROW, 2017. Ruim 5.700 nieuwe deelauto’s [WWW Document]. URL https://kpvvdashboard-4.blogspot.nl/ (accessed 12.15.17).
Lane, C., 2005. PhillyCarShare: First-Year Social and Mobility Impacts of Carsharing in Philadelphia, Pennsylvania. Transp. Res. Rec. J. Transp. Res. Board 1927, 158–166. https://doi.org/10.3141/1927-18
Le Vine, S., Polak, J., 2017. The impact of free-floating carsharing on car ownership: Early-stage findings from London. Transp. Policy. https://doi.org/10.1016/j.tranpol.2017.02.004
Martin, E., Shaheen, S., Lidicker, J., 2010. Impact of Carsharing on Household Vehicle Holdings. Transp. Res. Rec. J. Transp. Res. Board 2143, 150–158. https://doi.org/10.3141/2143-19
Meijkamp, R.G., 2000. Changing consumer behaviour through Eco-efficient Services - An empirical study on Car Sharing in the Netherlands. TU Delft.
Millard-Ball, A., Murray, G., Schure, J. Ter, Fox, C., Burkhardt, J., 2005. Car-Sharing: Where and How it Succeeds. Washington, D.C.
Mueller, J., Schmoeller, S., Giesel, F., 2015. Identifying Users and Use of (Electric-) Free-Floating Carsharing in Berlin and Munich. IEEE Conf. Intell. Transp. Syst. Proceedings, ITSC 2015–Octob, 2568–2573. https://doi.org/10.1109/ITSC.2015.413
Münzel, K., Boon, W., Frenken, K., Vaskelainen, T., 2017. Carsharing business models in Germany: characteristics, success and future prospects. Inf. Syst. E-bus. Manag. 1–21. https://doi.org/10.1007/s10257-017-0355-x
Nijland, H., van Meerkerk, J., 2017. Mobility and environmental impacts of car sharing in the Netherlands. Environ. Innov. Soc. Transitions. https://doi.org/10.1016/j.eist.2017.02.001
Prettenthaler, F.E., Steininger, K.W., 1999. From ownership to service use lifestyle: the potential of car sharing. Ecol. Econ. 28, 443–453. https://doi.org/10.1016/S0921-8009(98)00109-8
Prieto, M., Baltas, G., Stan, V., 2017. Car sharing adoption intention in urban areas: What are the key sociodemographic drivers? Transp. Res. Part A Policy Pract. 101, 218–227.
37
https://doi.org/10.1016/j.tra.2017.05.012
Rifkin, J., 2015. The Zero Marginal Cost Society : the internet of things, the collaborative commons, and the eclipse of capitialism. Palgrave Macmillan.
Rijksoverheid, 2015. Over drie jaar honderdduizend deelauto’s in Nederland.
Rogers, E.M., 2003. Diffusion of Innovation, 5th ed. Free Press, New York, NY, USA.
Shaheen, S., Cohen, A., 2007. Growth in Worldwide Carsharing: An International Comparison. Transp. Res. Rec. J. Transp. Res. Board 1992, 81–89. https://doi.org/10.3141/1992-10
Shaheen, S., Martin, E., Bansal, A., 2018. Peer-To-Peer (P2P) Carsharing: Understanding Early Markets, Social Dynamics, and Behavioral Impacts. https://doi.org/10.7922/G2FN14BD
Shaheen, S., Sperling, D., Wagner, C., 1998. Carsharing in Europe and North America: Past, Present, and Future. Transp. Q. 52, 35–52. https://doi.org/10.1068/a201285
Shaheen, S., Wipyewski, K., Rodier, C., Novick, L., 2004. Carlink II: a commuter carsharing pilot program final report. Transportation (Amst). 1–186.
Shaheen, S.A., 1999. Dynamics in Behavioral Adaptation to a Transportation Innovation: A Case Study of Carlink–A Smart Carsharing System California PATH Research Report. Berkeley.
Shaheen, S.A., Mallery, M.A., Kingsley, K.J., 2012. Personal vehicle sharing services in North America. Res. Transp. Bus. Manag. https://doi.org/10.1016/j.rtbm.2012.04.005
SmartAgent, 2011. Utrechts Autodelen : perceptie en praktijk.
Sperling, D., 2018. Three Revolutions. Steering Automated, Shared and Electric Vehicles to a Better Future. ISLAND PRESS.
Steer Davies Gleave, 2017a. Carplus annual survey of car clubs 2016/17 Scotland. Leeds.
Steer Davies Gleave, 2017b. Carplus Annual Survey of Car Clubs 2016/17 London. Leeds.
Steer Davies Gleave, 2016. Carplus Annual Survey of Car Clubs 2015/16 England and Wales (excluding London). Leeds.
Steininger, K., Vogl, C., Zettl, R., 1996. Car-sharing organizations. Transp. Policy 3, 177–185. https://doi.org/10.1016/S0967-070X(96)00024-8
Truffer, B., 2003. User-led Innovation Processes: The Development of Professional Car Sharing by Environmentally Concerned Citizens. Innov. Eur. J. Soc. Sci. Res. 16, 139–154. https://doi.org/10.1080/13511610304517
Wappelhorst, S., Graff, A., Steiner, J., Hinkeldein, D., 2013. New carsharing offers and customer groups: Implications for a growing and diversifying market, in: European Transport Conference.
Wilhelms, M.-P., Henkel, S., Merfeld, K., 2017a. You Are What You Share: Understanding Participation Motives in Peer-to-Peer Carsharing, in: Meyer, G., Shaheen, S. (Eds.),
38
Disrupting Mobility - Impacts of Sharing Economy and Innovative Transportation on Cities, Lecture Notes in Mobility. Springer International Publishing, Cham, pp. 105–120. https://doi.org/10.1007/978-3-319-51602-8
Wilhelms, M.-P., Merfeld, K., Henkel, S., 2017b. Yours, mine, and ours: A user-centric analysis of opportunities and challenges in peer-to-peer asset sharing. Bus. Horiz. https://doi.org/10.1016/j.bushor.2017.07.004
Wilhelms, M.P., Henkel, S., Falk, T., 2016. To earn is not enough: A means-end analysis to uncover peer-providers’ participation motives in peer-to-peer carsharing. Technol. Forecast. Soc. Change. https://doi.org/10.1016/j.techfore.2017.03.030
39
Appendix
Correlations 1 2 3 4 5 6 7 8 9 10 11 12
1
Adoption (Adopter, potential adopter, no adopter)Carsharer, potential carsharer or no carsharing
Pearson Correlation 1
N 1835
2
Car suppliers (P2P provider, Private provider, no provider but car owner)
Pearson Correlation .387** 1
N 1474 1474
3 Age Pearson Correlation .135** -0.011 1
N 1835 1474 1835
4 Gender (male=1, female=2)
Pearson Correlation 0.017 -0.034 -.148** 1
N 1835 1474 1835 1835
5 Income Pearson Correlation -0.011 0.055 .068** -.143** 1
N 1458 1160 1458 1458 1458
6 Education Pearson Correlation -.179** -.110** -.165** 0.010 .191** 1
N 1835 1474 1835 1835 1458 1835
7 Dummy G4 city Pearson Correlation -.149** -.059* -0.034 0.006 -0.021 .191** 1
N 1835 1474 1835 1835 1458 1835 1835
8 Dummy Children in household
Pearson Correlation -0.042 -0.020 -.310** .096** .152** .052* -.086** 1
N 1835 1474 1835 1835 1458 1835 1835 1835
9 Dummy No car in household
Pearson Correlation -.297** -.150** -.166** 0.045 -.268** .100** .297** -.146** 1
N 1835 1474 1835 1835 1458 1835 1835 1835 1835
10 Dummy Public transit subscripton
Pearson Correlation -.194** -0.043 0.045 0.010 -0.028 .121** .129** -.141** .270** 1
N 1826 1466 1826 1826 1451 1826 1826 1826 1826 1826
11 Dummy Green party voter
Pearson Correlation -.149** -.054* -.052* 0.047 -0.028 .144** .085** -0.020 .142** .100** 1
N 1645 1315 1645 1645 1299 1645 1645 1645 1645 1636 1645
12 Crime numbers per 1000 inhabitants
Pearson Correlation -.089** -.070** -0.018 0.024 -0.001 .124** .401** -0.040 .164** .074** 0.026 1
N 1832 1471 1832 1832 1456 1832 1832 1832 1832 1823 1643 1832
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
40
Only CS users 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 Use frequency low (1), high (2)
Pearson Correlation 1 N 284
2
Would have bought a car (1), maybe (2), not (3) if had not started carsharing
Pearson Correlation 0.038 1
N 284 284
3 P2P user Pearson Correlation -0.107 -0.076 1 N 284 284 284
4 B2C user Pearson Correlation .325** 0.052 -.438** 1 N 284 284 284 284
5 Age Pearson Correlation -0.009 0.002 -0.018 -0.068 1 N 284 284 284 284 284
6 Gender (male=1, female=2)
Pearson Correlation -0.021 0.048 0.085 -0.007 -.176** 1 N 284 284 284 284 284 284
7 Income Pearson Correlation 0.087 -.136* -0.120 .175** 0.094 -.179** 1 N 238 238 238 238 238 238 238
8 Education Pearson Correlation .175** -0.003 -0.030 .117* -.145* 0.080 .219** 1 N 284 284 284 284 284 284 238 284
9 Dummy G4 city Pearson Correlation 0.012 0.110 -.207** 0.030 -0.036 -0.006 0.071 0.084 1 N 284 284 284 284 284 284 238 284 284
10 Dummy Children in household
Pearson Correlation 0.088 -0.038 0.087 -0.017 -.182** 0.090 0.063 0.050 -0.085 1 N 284 284 284 284 284 284 238 284 284 284
11 Dummy No car in household
Pearson Correlation .163** .426** -.166** 0.111 -0.056 0.056 -.144* 0.103 .247** -.157** 1 N 284 284 284 284 284 284 238 284 284 284 284
12 Dummy Public transit subscripton
Pearson Correlation .241** .170** -0.024 .238** 0.066 -0.009 0.112 0.019 0.024 -.119* .278** 1 N 283 283 283 283 283 283 237 283 283 283 283 283
13 Dummy Green party voter
Pearson Correlation 0.055 0.082 0.031 -0.006 0.037 0.013 -0.026 .217** -0.012 -0.047 .194** 0.116 1 N 255 255 255 255 255 255 213 255 255 255 255 254 255
14 Most important reason to carshare: Costs
Pearson Correlation -0.017 .304** -0.050 0.065 -0.050 0.023 -0.103 0.045 0.104 -0.045 .451** .162** 0.070 1 N 284 284 284 284 284 284 238 284 284 284 284 283 255 284
15 Most important reason to carshare: Convenience
Pearson Correlation 0.086 0.106 -0.107 0.039 -0.060 0.009 -0.030 -0.012 -0.039 -0.025 0.088 -0.010 -0.024 -.272** 1
N 284 284 284 284 284 284 238 284 284 284 284 283 255 284 284
16 Most important reason to carshare: Environment
Pearson Correlation 0.041 0.007 0.081 -0.072 .167** 0.082 0.044 -0.020 -.167** 0.016 -.196** -0.026 0.059 -.262** -0.109 1
N 284 284 284 284 284 284 238 284 284 284 284 283 255 284 284 284
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).