Transportation

https://doi.org/10.1007/s11116-024-10514-4

Psychological factors impacts on carsharing use

Mohamed Abouelela1 · Christelle Al Haddad1 · Constantinos Antoniou1

Accepted: 11 July 2024

© The Author(s) 2024

Abstract
Carsharing services have a significant potential for improving urban mobility by increasing
the independence and freedom of travel and reducing traffic externalities. Although car-
sharing has been used for over a decade, several aspects need further investigation, such as
the impact of user’s psychological factors on service use, as well as the factors impacting
users’ choices between different carsharing operators, in particular their preferences for dif-
ferent payment schemes, and their perceptions of the operators’ application rating. Accord-
ingly, four hybrid choice models (HCM) were estimated to investigate factors impacting
(i) the knowledge about carsharing services, (ii) carsharing adoption, (iii) the shift from
other modes to carsharing, (iv) the choice between carsharing operators with different pay-
ment schemes, using a large survey sample (N = 1044 responses 9469 SP observation)
from Munich, Germany. The models showed the significance of sociodemographics, such
as income level, education level, household size, employment status, ownership of a bike,
access to a car, the availability of a driving license, and public transport subscription-based
tickets on the carsharing use directly and indirectly, and four psychological factors encom-
passing different personality traits (i.e., adventurous), travel behavior, and attitudes were
found to be significant in the various models; the latter covered service-related attitudes
(perceived carsharing app importance) and travel behavior attitudes or profiles (frequent
public transport user and frequent shared micromobility user). This research raises ques-
tions regarding the inequitable use of carsharing, the impacts of mobile applications on
using the service, and the potential of integrating carsharing in mobility as a Service plat-
forms to increase the potential for multimodality.

Keywords Shared mobility · Carsharing · Choice model · Attitudes · Personality

Introduction

In the past ten years, there has been a significant increase in the acceptance, utilization,
development, and improvement of app-based shared mobility services. This growth has
been made possible by revolutionary advancements in information and communication

* Mohamed Abouelela
[email protected]
1
Transportation Systems Engineering, Technical University of Munich, Arcisstrasse 21, Munich,
Germany

Vol.:(0123456789)
 Transportation

technologies (ICT). Shared mobility services encompass various options, including

schemes, services, vehicles, and business models. Examples of these services include ride-
sharing and carpooling, in which people share or split a ride, as well as carsharing and
shared micromobility options such as bikesharing and e-scooter-sharing, in which vehicles
can be rented based on time or distance (Narayanan and Antoniou 2022; Gilibert and Ribas
2019). These new mobility services have changed the landscape of urban mobility by intro-
ducing the concepts of on-demand services and pay-per-use, which increases the attractive-
ness of such services for users due to their ease of use, ease of payment, convenience, but
also as they are perceived to be convenient, safe, and environmentally friendly (Arteaga-
Sánchez et al. 2020; Tirachini and del Río 2019; Watanabe et al. 2017; Rayle et al. 2016).
Shared mobility services benefits are not limited to the individual level but it could
extended to the city scale and could be an attractive solution for various transportation
problems as they do not need large infrastructure investments and are quick to imple-
ment in most of the cases (Abouelela et al. 2022). Maintaining, upgrading, and construct-
ing transportation infrastructure generally needs significant investments and a long time
to materialize, which is not always a viable solution; one example is extending the trans-
portation system’s accessibility to suburban areas with inefficient public transportation’s
access (Burghard and Dütschke 2019; Abouelela et al. 2022). Shared mobility services
could reduce the demand and congestion on roads, as well as the vehicle kilometer trave-
led (VKT), under the conditions of not replacing public transportation trips and replacing
low occupancy vehicles (Tirachini et al. 2020). Alonso-Mora et al. (2017) concluded that
shared rides could as well reduce the number of cars on city roads. The same promises of
reducing the number of vehicles on the streets could be achieved using carsharing services,
as private cars are idly parked for around 90% of the time (Zhang et al. 2015). Transport for
London (TfL) sees carsharing services as complementary to public transportation services
(Akyelken et al. 2018). Carsharing use could even be correlated with the increase in public
transportation use (Aguilera-García et al. 2022).
On the other hand, challenges arise with the introduction of shared mobility should not
be overlooked. Several problems have been raised since the introduction, such as increased
safety concerns, and the rapid increase of numbers of reported severe and fatal accidents
related to shared scooter use (Yang et al. 2020). Other problems related to operational-
aspects and governance were also noted, including but not limited to, fleet-size control cap-
ping and organization, and permit cost (Gössling 2020). In most of the cases the evaluation
of shared mobility impacts on CO2 emissions did not consider the whole life cycle assess-
ment, and when the whole life cycle was assessed for the cases of shared e-scooter, the
results showed that scooters are producing CO2-equivalent per passenger-kilometer more
than the modes they replace (Moreau et al. 2020), in addition to the other negative impacts
that could result from shared services, such as attracting commuters from other sustainable
modes, and generating extra kilometers traveled (VKT) during redistribution and mainte-
nance processes (Møller and Simlett 2020). Also, the use of shared mobility require access
to the Internet, smart–mobile–phone, and digital banking systems, which is not always
available for all the population groups, and thus, it creates equitable use problem by being
mostly accessible by the high income groups in comparison to the rest of the population
(Abouelela et al. 2024).
Overall, the system is attractive to implement as establishing its infrastructure is con-
sidered relatively quick, its market is growing and it has the potential to continue growing
in the future (Morwinsky 2023; Shaheen and Cohen 2020); however, it is not always an
easy process as these services could face several challenges, such as economic viability
(Golalikhani et al. 2021; Poltimäe et al. 2022), in addition to other obstacles, such as use
Transportation

complexity, limited availability leading to unreliability of the service, responsibility for the
usage, long access time, low public awareness, and the required higher population density
for economically feasible operation (Nansubuga and Kowalkowski 2021). An example of
the struggle is the ShareNow1 cease of operations in North America and several European
cities, such as London, Brussels, and Florence in 2020 (Miljure and Ben 2019). Share-
Now attributed ending the service to the rapid changes in the urban mobility landscape, the
intense competition with different providers of different services, the unavailability of ade-
quate infrastructure (e.g., charging infrastructure for electric vehicles), and the increased
operation cost raising concerns regarding the long term economic model feasibility (Pyzyk
2019; Miljure and Ben 2019). On the other hand, there is a recent increase in the service’s
popularity and adoption in Western European cities and Russia. This increase is condi-
tional based on the city’s characteristics, such as the population’s educational level, uni-
versity presence, and the number of Green party voters (Kireeva et al. 2021; Münzel et al.
2020). Therefore, it is essential to note that the implementation of the service and its eco-
nomic viability is not always granted, and careful consideration, especially in the planning
phase, should be attained.
Carsharing is a form of shared mobility that provides easy access to on-demand car use
without the burden of car ownership responsibilities, the need to process paperwork such
as for car rental services, or even the need to return the vehicle to the pickup points in most
of the cases (Liao and Correia 2022). Carsharing services operate on different schemes
when it comes to pickup and drop-off arrangements, with three main schemes: round-trip
systems, where users pick up and return vehicles to the same locations, and more flexible
options like one-way trips or free-floating systems, where vehicles can be dropped off at
any designated point within a specified area (Amirnazmiafshar and Diana 2022; Jorge and
Correia 2013).
Carsharing services and other shared mobility services are not only changing the land-
scape of urban mobility, but also the traditional idea of a car manufacturer producing, buy-
ing, and selling vehicles. Currently, some leading car manufacturers are promoting them-
selves as mobility providers (Akyelken et al. 2018), including Mercedes–Benz Group,
BMW, Volkswagen, Toyota, and General Motors. Mercedes-Benz Group has two carshar-
ing services (ShareNow, and Croove), acquired two taxi services (myTaxi,2 and Hailo), is
investing in two ride-hailing services (Via,3 and Blacklane4), and starting its own mobility
platform moovel5 (Akyelken et al. 2018). Therefore, there is an essential need to under-
stand in–depth the different aspects of these services for better operation and integration
within the urban environment.
Some of the main aspects of shared mobility that are important for the different stake-
holders are the socio-demographic characteristics of the users and their general travel
behavior, as well as their impacts in deriving the demand and identifying user target groups
(Jochem et al. 2020). Psychological factors such as attitudes, perceptions, and personality
traits play a significant role in individual travel behavior and mode choices (Kroesen and
Chorus 2020). The importance of understanding the impact of psychological factors on

1
ShareNow is a merger company between Car2Go the carsharing subsidiary of BMW, and the DriveNow,
the subsidiary of Mercedes-Benz Group (https://​share-​now.​com).
2
https://​free-​now.​com, now the service is a joint venture between Mercedes-Benz Group and BMW.
3
https://​info.​ridew​ithvia.​com.
4
https://​black​lane.​com.
5
https://​moove​lus.​com, the platform is one of the Mobility as a Service (MaaS) providers.
 Transportation

travel behavior and mode choice lies in their ability to facilitate encouraging the use of the
modes of interest, as they could be described as the underlying motivation for specific mode
use (Bhagat-Conway et al. 2024). Previous research has shown that attitudes were found to
have a significantly higher impact on the use of shared mobility as compared to sociode-
mographics, such as in the case of pooled rides (Abouelela et al. 2022). Still, there is a gap
in terms of existing research on attitudes and personality traits in the scope of carsharing
and shared mobility in general, mostly when comparing it to studies focusing on sociode-
mographics, which have been well examined and explored in the literature (Monteiro et al.
2023; Efthymiou and Antoniou 2016; Efthymiou et al. 2013). Several of these psychological
factors are still under exploration and their role in the mode choice travel decision (Rahimi
et al. 2020a) in general, and shared mobility use in particular, is not well understood.To
the best of the authors’ knowledge, many aspects of carsharing services have not yet been
studied, such as the perceived service and feature offerings by different carsharing operators,
including digital operator aspects (often reflected in the operator rating on the app store), as
well as their impact on service adoption and use frequency (Monteiro et al. 2022).
The digital dimension of the carsharing services has also not been investigated in-
depth, and includes the mobile application friendliness and ease of use, the service pro-
vider’s website landing page, the digital marketing of the service, the online marketing
campaigns, and the business-to-business offers (Janasz and Schneidewind 2017). Another
service feature to consider is the operator payment schemes (per minute or kilometer as
recently introduced by some operators). The impact of the above-mentioned features on
the operator choice still needs to be investigated. Finally, carsharing research on adop-
tion and use has not yet been totally understood due to the novelty of the services; a
large number of the carsharing studies have been completed before the services were
even launched or during the early operational and adoption stages, during which users
might have a different use behavior as they are getting familiar with the service. Another
motivation of this paper is therefore to contribute to the existing body of research with
more timely study in which the operation of carsharing services is ongoing at the time
the research is done (Le Vine and Polak 2019; Hjorteset and Böcker 2020).
We therefore contribute to the current literature by updating the knowledge regarding
carsharing use, using user-level information through a large online survey, and answering
the following two research questions (RQ) investigating the roles of users’ psychological
factors: personal attitudes, travel behaviour, and carsharing-related features on the different
aspects of carsharing services.

• RQ1) How do users’ psychological factors impact carsharing adoption and use?
• RQ2) What factors impact the choice between different carsharing operators?

The rest of the article is organized as follows; “Literature review” section summarizes
some of the selected studies related to user factors and attitudes impacting carsharing use
and the different service-related characteristics that impact user’s choices. In “Methods and
study set-up” section explains the methods used in the research and the case study setup
used for the analysis and modeling. In “Analysis results” section spans across two parts that
answer the research questions (RQ1 and RQ2); first, we analyze the collected data, second,
we model the different factors that impact carsharing adoption and use, with a special focus
on personality traits and attitudes. We also model and extract the factors impacting users’
choices between different carsharing operators. Finally, “Discussion, limitations, and con-
clusions” section discusses the study findings, highlights the policy implications, and sum-
marizes the conclusion.
Transportation

Literature review

The benefits of carsharing

Sustainability is one of the many benefits associated with carsharing; it is considered a

sustainable mode of transportation that has a wide array of positive impacts on the urban
environment, such as the reduction in household car ownership and, subsequently a reduc-
tion in Greenhouse Gas (GHG) emissions that could reach up to 30–54% as a consequence
of reduced Vehicle Miles/Kilometers Travelled (VMT/VKT) (Shaheen et al. 2019; Nijland
and van Meerkerk 2017; Martin and Shaheen 2011a). Also, electrification of the carshar-
ing fleet was proven to be environmentally advantageous (Luna et al. 2020) and was able
to yield more than 30% reduction on carsharing users’ GHG even if there was no change in
VKT (Namazu and Dowlatabadi 2015). Several examples of the previous positive poten-
tials of carsharing use were observed; in Germany, evidence associated with the reduction
of car ownership and the number of station-based carsharing in the same area were present
(Kolleck 2021); in China, in 2017, carsharing has caused a significant reduction of energy
used, and CO2 emissions, with the expectation of higher savings by 2025 (Te and Lianghua
2020).
Shaheen et al. (2019) and Martin and Shaheen (2016) observed a decline in the average
VKT of carsharing users ranging from 6 to 63% in North America, considering several
conditions such as giving up car ownership, and the type of the service one-way or round
trip. These tendencies were further corroborated by studies in Palermo, Italy (Migliore
et al. 2020), the Netherlands (Nijland and van Meerkerk 2017), and London, UK (Wu et al.
2020). Also, Wu et al. (2020) noted that in London, higher satisfaction with the proximity
to carsharing vehicles contributed to a larger reduction in VKT. Interestingly, carsharing
users who live in suburban areas tend to drive fewer kilometers than their counterparts
in dense urban areas (Clewlow 2016), which could be attributed to the lower density of
available vehicles in the suburbs, resulting in carsharing users canceling the non-essential
trips. Other environmental-related positive impacts, such as saving materials and reducing
wastes, were observed (Harris et al. 2021). The environmental impacts of carsharing and
their total magnitudes are heavily dependent on the occupancy rate, the used vehicle and
fuel type of the fleet, the modal share of carsharing, the modes replaced by carsharing, and
the vehicle’s lifespan (Poltimäe et al. 2022; Harris et al. 2021; Jung and Koo 2018). Also,
adverse environmental impacts of carsharing were observed; in Palermo, Italy, the fleet
only contains diesel and natural gas vehicles, and it was found to increase the CH4 and NOx
emissions of the city (Migliore et al. 2020).
Impacts related to infrastructure and built-environment were also observed, as carshar-
ing use can help in reducing car ownership rates, which promotes more positive impacts
on curb-side management, minimizing the space uptake for car parking (Golalikhani et al.
2021). A study among students in Ithaca, New York by Stasko et al. (2013) found that
since the introduction of carsharing in the area, student parking permit sales had declined
despite a continuous increase in enrollment. The causality of these occurrences was not
investigated or verified. Another survey in France investigated carsharing use impact on
parking derived that for every carsharing vehicle on the street, between 1.6 and 4.2 on-
street parking spaces, 0.3–0.6 public parking spaces, and 2.1–4.2 private parking spaces
could be eliminated (6t-Bureau de recherche and ADEME 2016). Diana and Chicco (2022)
analyzed the spatial distribution of changes in parking demand related to carsharing and
 Transportation

found that more relieved parking spaces could be anticipated for central areas, while more
negative impacts might be imposed on the parking in peripheral areas.
Integrating carsharing with public transportation would yield more benefits by extend-
ing the spatiotemporal accessibility of public transportation. Some examples of this inte-
gration are the decentralized mobility hub (Czarnetzki and Siek 2022), implementation
of dedicated carsharing facilities (Engel-Yan and Passmore 2013) and unbundled parking
(Schure et al. 2012) in residential buildings, and appropriate financial and policy back-
ing from the authorities in forms of aids to the low income-groups (Rabbitt and Ghosh
2013; Bocken et al. 2020). Note that the extent of carsharing impacts could highly vary
depending on the region, built environment (Clewlow 2016; Jain et al. 2022), accessibil-
ity of public transportation, and the carsharing replaced modes (Shaheen et al. 2019; Jain
et al. 2022; Kolleck 2021; Duncan 2011). Other positive potentials for carsharing use were
observed, but they were less explored, such as benefits associated with the B2B carsharing
model capabilities of reducing work trip cost as the car can be used without bearing own-
ership-related costs and duties, increasing thereby not only trip sustainability, but also the
workplace attractiveness, which could now subsidize carsharing trips for their employees
(von Wieding et al. 2022). Carsharing trips were also found to encourage multimodality,
physical activities, and a healthier lifestyle (Kent 2014; Shaheen et al. 2019; Harris et al.
2021). Also, carsharing was found to increase access to cars for car-less households (who
do not own private vehicles), providing them thereby with more independence and equi-
table access to opportunities (Stasko et al. 2013; Kent 2014; Shaheen et al. 2019). This in
turn improves the mobility of lower-income groups by increasing the number of available
travel options (Kumar Mitra 2021), and strengthening the sense of community among users
(Hartl and Hofmann 2022; Harris et al. 2021).

Factors impacting the adoption and use of carsharing

Several factors impact the adoption and use of carsharing services; these factors could be
categorized into three main groups; (i) service-related factors, (ii) exogenous factors, and
(iii) user-related factors. The first group of factors included the number of available vehi-
cles in the stations, and vehicle age; in a study by De Lorimier and El-Geneidy (2013),
this encouraged carsharing use in Quebec, Canada. In metropolitan Vancouver, lowering
the membership fees was found to attract more users (Namazu et al. 2018). The difference
between the trip cost, and the mode carsharing replaced was found to be the most signifi-
cant factor impacting carsharing use in Beijing, China (Yoon et al. 2017). Personalized use
incentives were also found to attract more users (Feng et al. 2023). In Shanghai city, elec-
trical vehicle battery charging level and the number of available vehicles in stations impact
the user choice for the vehicles (Hu et al. 2018b). Secondly, exogenous factors are also key
such as adverse weather conditions (Yoon et al. 2017), availability, accessibility of public
transportation station (Balac et al. 2015; Hu et al. 2018a; Khan and Machemehl 2017),
land-use (Kim et al. 2012; Stillwater et al. 2009), intersection and road density, and the
availability of parking (Chen wt al. 2018; Yoon et al. 2017; Hu et al. 2018a).
Thirdly, several studies focused on investigating the sociodemographic characteris-
tics influencing carsharing use. The findings of these studies have identified users as
young, male, well-educated, with high-income, and full-time employment (Le Vine and
Polak 2019; Martin and Shaheen 2011a; Alemi et al. 2018; Ahmed et al. 2021; Luo et al.
Transportation

2019). The role of other important personal drivers to the service is less known, and
here we mean the personal attitudes and personality traits, despite the fact that there is
evidence suggesting the significance of attitudes on the use and adoption of carsharing
services, noting that understanding personal attitudes is claimed to enhance the models
predictability (Pronello and Gaborieau 2018). For instance, carsharing users are more
likely to own “greener” vehicles (Clewlow 2016) and exhibit more eco-friendly behav-
ior (Jung and Koo 2018), hinting at higher concerns towards environmental issues, and
carsharing advocacy attitude increased the adoption of carsharing compared to other
modes (Li and Kamargianni 2020). In the realm of carsharing, research on the role of
personal attitudes has yielded mixed conclusions.
Zhang and Li (2020) and Li and Zhang (2021) discovered that subjective/social
norms had the biggest influence on the intention to use carsharing, and attitudinal vari-
ables, including environmental concerns, imposed a much more limited impact, while
a study in Taiwan (Buschmann et al. 2020) reported the complete opposite. Varieties
also exist within the range of behavioral constructs that were found to be significant in
carsharing familiarity and adoption. Aguilera-García et al. (2022) found that high shar-
ing propensity, variety-seeking lifestyle, and preference for driving positively impacted
familiarity, and that pro-environmental behaviors reduced carsharing usage. On the
other hand, Thurner et al. (2022) concluded that people who were believers of science
and technology, who were generally early adopters of novel technology, and those with
self-expressive social values tended to be carsharing adopters. The previous discrep-
ancies are unsurprising, considering the virtually unlimited spectrum of attitudes that
humans might have. Yet, researchers are constrained to investigate only a select few,
along with behavioral indicators which vary across the board.
Cultural context also moderates the effects of other sociodemographic variables. For
instance, society could be more concerned about conforming to the norms than their
expressions, leading to subjective norms being more influential in their decisions. Also,
there is a complex interrelation between these attitudinal constructs, which is hard to
interpret. This is well demonstrated by Zhang and Li (2020), Burghard and Scherrer
(2022), Li and Zhang (2021), Acheampong and Siiba (2020) in which environmental
attitudes imposed no direct impact or even negative impact on carsharing intentions,
while simultaneously being positively correlated with another construct which in
turn positively impacted the carsharing intention (i.e., positive indirect impact). This
shows how the role of attitudes in human decision is a complex topic and requires fur-
ther research with a possibly wider range of attitude constructs. For example, Hjorte-
set and Böcker (2020) further differentiated the resulting attitude towards carsharing
into general interest, anticipated intention, and actual decision to utilize the service.
Another part of human attitudes is personality traits, which are the main drivers of
travel demand (Mokhtarian et al. 2001). Different personality traits are hypothesized to
impact travel behavior differently; while the adventure-seeker personality was found to
be likely to travel and drive faster than other personalities, are prone to have and create
more elements of danger (Furnham and Saipe 1993). Redmond (2000) concluded that
people with adventure-seeking personalities are more likely to enjoy leisure trips over
work trips and may also prioritize them. Another personality associated with the prefer-
ence for using private cars over public transport is the organizer personality (Redmond
2000). A summary of the factors impacting carsharing use is presented in Fig. 1 below.
 Transportation

Variety
seeker
Infrastructure
Efficient
Land use
Availability Sharing
of other propen-
sity Personalty
Science
modes Preference
believer
traits
Adventure
for
driving seeker

Exogenous Advocacy
of car-
Weather factors sharing Personal
conditions attitudes Car
ownership
Env.
friendly

Income
Factors im-
pacting car- User-related
sharing use
Trip cost
Employment
status

Membership Age Gender

fees Service-
related Personalized
incentives

Battery
charging
level
Vehicle age
Number
of vehicles
in station

Fig. 1  Summary of factors impacting carsharing adoption and use (own illustration)

Synergies between carsharing use and travel behavior

Carsharing might play a significant impact on travel behavior and users’ long and short-
term travel decisions. It can impact the decision to give-up a car and forego/delay the deci-
sion to acquire a new one (Ko et al. 2019; Seo and Lee 2021). Although varying conclu-
sions exist across case studies, the general consensus suggests a decline in the level of car
ownership, with studies quoting four (Migliore et al. 2020; Shaheen et al. 2018) to twenty-
three (Lane 2005) private vehicles being replaced for every carsharing vehicle in operation.
This conclusion is consistent with Le Vine and Polak (2019) findings, which highlighted,
based on a survey in London (N = 347 responses), that as much as 37% of respondents
had their car ownership decisions impacted by using carsharing, as users opted to drop the
decision of buying a car or dispose of their currently owned car. Factors affecting a user to
dispose or forego buying a car include income level, age, housing type, satisfaction towards
the carsharing service, access time to carsharing station, fuel type, and the price or cost of
the service (Jung and Koo 2018; Ko et al. 2019). Simultaneity bias can also be a concern
as Jain et al. (2020) found within their case study; carsharing mostly acted as an enabler of
Transportation

mobility lifestyle change but was not the primary cause of households shedding their pri-
vate cars, as life events had a stronger influence.
Furthermore, carsharing’s overall impact on sustainability depends on the modes it
replaces, whether they are “greener” and more active modes such as walking, cycling, and
public transportation (Chicco and Diana 2021). The impact of carsharing on the general
car use is less conclusive as some studies reported that the majority of carsharing users
drove less frequently (than before carsharing adoption) (Martin and Shaheen 2011b;
Shaheen et al. 2018), while others studies claimed the contrary (Stasko et al. 2013; Martin
and Shaheen 2016). This is due to the fact that the effect of those who dispose of private
cars is counterbalanced by the impact of those who gain access to cars through carsharing
(Lane 2005). While such studies often relied on user surveys, the latter have often been
criticized as they focus on carsharing users and therefore create self-selection biases, which
might impact the conclusions.

Modeling techniques

Attitudes are often treated as latent variables derived from stated behavioral statements. To
capture these latent attitudes and determine the indicating constructs, several methods have
been used in the past, including Structural Equation Models (SEMs) (Yazdanpanah and
Hosseinlou 2016; Aguilera-García et al. 2022; Rahimi et al. 2020b; Zhang and Li 2020),
Principal Component Analysis (PCA) (Queiroz et al. 2020; Thurner et al. 2022), and Latent
Class Analysis (LCA) (Olaru et al. 2021). Subsequent regression analysis [e.g., Bivariate
Logit (Queiroz et al. 2020), and Hybrid Choice Model (HCM), or Integrated Choice and
Latent Variable models (ICLV) (Sun et al. 2021)] incorporating the latent attitudinal vari-
ables in models is frequently conducted to assess the causality between attitudes and other
variables in question (e.g., acceptance of carsharing). The main objective of this integra-
tion is to enhance the model’s ability to understand the choice process by incorporating the
user’s cognitive behavior, attitude, and psychological factors into the choice model. This
integration also aims to improve the model’s goodness of fit where applicable (Vij and
Walker 2016; Temme et al. 2007; Ben-Akiva et al. 1999). ICLV models were, for instance,
used to quantify the factors impacting the frequency of pooled-rides uses in Mexico City,
Mexico (Abouelela et al. 2022). Theory of Planned Behavior (Jain et al. 2021; Zhang and
Li 2020; Li and Zhang 2021), Rogers et al. (2014)’s Theory of Innovation Diffusion (Jain
et al. 2021; Burghard and Scherrer 2022), and the Theory of Reasoned Action (TRA) along
with its extensions (Buschmann et al. 2020) are often incorporated in assessing the role of
personal attitudes. Further scientific frameworks that are prevalent in this research topic
are the Technology Acceptance Model (TAM) (Al Haddad et al. 2020; Schlüter and Weyer
2019; Buschmann et al. 2020) and its modifications, such as the Unified Theory of Accept-
ance and Use of Technology (UTAUT) (Fleury et al. 2017).

Gap analysis

The review of the current research shows several gaps in the current body of literature
that need to be bridged to utilize the maximum benefits of shared mobility use, specifi-
cally carsharing services. First of all, a significant portion of carsharing-related studies
were developed before the implementation of the services or during the early deployment
stages. Accordingly, there is a pressing need to update the current literature with more
recent studies, especially for users who are already familiar with the service and have used
 Transportation

it for an extended period (Hjorteset and Böcker 2020; Le Vine and Polak 2019; Namazu
et al. 2018). The importance of covering this point is to test the previously anticipated or
observed pre-and-early-use behavior of carsharing and decide if the current operations and
policies might need modification or changes.
Secondly, the investigating of the impact of carsharing-related features, such as opera-
tor offering and the used mobile app, on service use, adoption, and choice between differ-
ent operators is still scarce in the existing body of the literature and generally overlooked
(Monteiro et al. 2022). As discussed in the following sections, such aspects significantly
impact service adoption and use.
Finally, the impacts of psychological factors such as personal attitudes and personality
traits on the use and adoption of carsharing services are not well established (Aguilera-
García et al. 2022), despite their essential impacts on travel behavior and mode choice, and
the use of the different shared mobility options (Abouelela et al. 2022). Understanding the
interaction between the different psychological factors and travel behavior is essential to
understanding how they could alter the current travel behavior to be more sustainable (Steg
2007). In other words, shared mobility, especially carsharing services, a niche service, and
understanding these psychological factors impacting their use could help in driving poli-
cies that increase the sustainable adoption of these services (Burghard and Scherrer 2022),
for example, social norms were found to be positively increasing the adoption of carsharing
when it was promoted as a low-carbon mobility option in Sweden (Mundaca et al. 2022b,
a). Also, the role of psychological factors is not only limited to understanding and altering
travel behavior, but it also could help in anticipating which policies, such as environmental
policies, could be accepted (Ejelöv and Nilsson 2020). Therefore, we are trying to under-
stand additional psychological factors that might impact carsharing use, to understand how
to integrate the service with other modes of transport, and which factors could attract users
to the service or be used as a pull measure.
In summary, this study addresses the gaps mentioned above by testing the impacts of
different psychological factors, namely personality traits, regular travel behavior, and atti-
tudes (importance of service-related features), on carsharing use and adoption, thereby
answering the research questions.

Methods and study set‑up

Methods

Survey design

The main goal of this research is to understand the impacts of attitudes, travel behavior,
and personality traits on the use of carsharing services; therefore, we designed a survey
in four parts, which was implemented online using Limesurvey platform (https://​Limes​
urvey.​com), and disseminated to different users group in Munich, Germany, during the
period of 20 of January to 25 of March 2022. We opted to deploy the survey online as it
was deployed during the COVID-19 pandemic, and we wanted to eliminate the chances
of infection during the data collection process. As carsharing users are likely young,
we targeted them in our data collection process. Young users are commonly adopters
of shared mobility in general and carsharing in particular, as highlighted in studies in
different locations, such as in Munich and Madrid (Aguilera-García et al. 2022), in
Transportation

Vancouver, Canada (Namazu et al. 2018), in Puget Sound region in the state of Wash-
ington, USA (Dias et al. 2017), and all over Germany (Burghard and Dütschke 2019).
Also, we collected data from non-users to check the different reasons for not adopt-
ing the service, as well as to evaluate the differences between the two groups. Overall,
we collected 1170 completed responses, and the average survey completion time was
12 min. The survey consisted of four main parts;

• In the first part, general travel behavior was investigated, where users were asked to
specify their usage frequency for different urban modes of transport, whether they
had a public transportation subscription ticket (such as a monthly ticket), whether
they owned bikes, e-bikes, a private car, and whether or not they had a valid driver’s
license in Germany. The modes that their use frequency was investigated are:

1. Bus 5. Personal bike 9. Taxi

2. Car as a passenger 6. Shared bike/E-bike 10. Train
3. Car as a driver 7. Shared E-scooter 11. Underground
metro
4. E-hailing 8. Suburban train 12. Walking

• In the second part, we investigated user familiarity with and usage of carsharing
services; we focused on usage frequency, willingness to walk to the vehicle pickup
location, trip purpose. Respondents were also asked about the modes they would
have used instead of carsharing for their last carsharing trip. Finally, respondents
were asked to evaluate the importance of different aspects of carsharing services,
such as mobile-application rating on the digital store, application ease of use, ser-
vice availability in different cities, service availability in EV, service availability in
the airport, service availability in different size vehicles (SUV, trucks, etc.), and the
availability of offers bundles (discounts, e.g., for all-day rental, and long-distance
rentals).
• The third part of the survey was the stated preference experiment; refer to Fig. 2. In this
experiment, respondents had to choose one carsharing service to perform an 11-kilo-
meter trip; the choice was between operator A, where the user pays a fixed cost per
kilometer. The other choice was operator B, where the trip cost would vary between a
minimum cost, an average cost, and a maximum cost based on congestion conditions.
The latter (cost range) would vary based on speeds (maximum, average, and minimum,
respectively) of previous trips (previous trip distribution).

Table 1 shows the attributes and their corresponding levels that were used for the exper-
iment. We opted to use travel cost, as it is a decisive factor in travel mode choice, and we
wanted to investigate two new factors that were not investigated previously, which are the
access distance users needed to walk to the nearest available vehicle and the service rating
on the digital application store. The attribute levels were calculated as follows:

Travel cost:
∗ Operator A, payment by km scheme, the average cost per km is 0.89 €/km,
obtained from the operator’s online website and is similar to values used by
Abouelela et al. (2021). A variation of this level (− 0.25%, 0%, +25%) would result
in a range of (0.66, 0.89, and 1.11) €/km.
 Transportation

Fig. 2  Scenario details and one block example

Table 1  Stated preference attributes and levels

Variable (unit) Levels
Operator A (payment by km) Operator B (payment by minutes)

Travel cost € [7.3, 9.8, 12.2] Minimum cost [5.6, 7.1, 9.2]
Average cost [8.1, 10.3, 13.2]
Maximum cost [12.1, 15.4, 19.8]
Access distance (m) [50, 100, 150] [50, 100, 150]
Application rating (⋆) [3, 4, 5] [3, 4, 5]
Engine type: electric Yes/No Yes/No

∗ Operator B, based on actual carsharing speed distribution. Essentially, minimum,

average, and maximum costs were calculated using the same carsharing trips in
Munich, Germany, for 2016, as described in Abouelela et al. (2021). The trip cost
was calculated based on the speed distribution and multiplied by the cost per min-
ute. The minimum cost was calculated based on the average speed for the first speed
quartile distribution. The average speed was calculated based on the average speed,
and the maximum cost was calculated based on the third-speed quartile average
speed. For each speed, subsequent cost (− 0.25%, 0%, +25%) values were calculated.
Transportation

∗ Operator B, costs per minute were obtained from operators’ online websites and
similar to the values used by Abouelela et al. (2021).
∙ The levels of access distance calculated for this experience considered that the walking
speeds are around 4–6 km/h (Bohannon and Andrews 2011), and that more than 50%
of pooled ride users opted to walk less than ten minutes for the ride pick up location
(Abouelela et al. 2022).
∙ Application rating on the digital application store was created specifically for this
experiment, as no similar attributes were not investigated before.
∙ Engine type was used to check the impact of the electric engine type on the user’s
choice, and it was a binary attribute with two levels: yes, and no. A similar attribute was
used by Monteiro et al. (2022).

The fourth part of the survey investigated the sociodemographic characteristics of the
users, where we asked users to specify their age, gender, education level, occupation, num-
ber of people and children in the household, and average monthly income. Also, in this
part, we asked users to specify their agreement on a five-points-scale (totally disagree,
disagree, neutral, agree, totally agree) on how much they identify with each of the 18 per-
sonality traits below, as used by Queiroz et al. (2020), Mokhtarian et al. (2001), Redmond
(2000):

1. Optimist 7. Like to stay close to home 13. Creative

2. Adventurous 8. Efficient 14. Calm
3. Like routines 9. Variety seeking 15. Anxious
4. Spontaneous 10. Punctual 16. Like being in
charge
5. Like being outdoor 11. Like to be alone 17. Participating
6. Risk taker 12. Independent 18. Lazy

Modeling framework

The main target of this research is to model the impact of attitudes and personality traits on
carsharing use, using the collected survey data. The survey consists of answers to attitudi-
nal and personality evaluation, revealed preference, and stated preference questions. The
different parts of the survey were used to answer the research question related to investigat-
ing factors impacting adoption, the shift from other modes, the choice between operators,
and finally, the knowledge or awareness level regarding carsharing service (essentially the
research questions RQ1 and RQ2). In investigating the examined factors, Hybrid Choice
Models (HCM) were estimated. The main purpose of estimating HCM models was to inte-
grate and investigate the impacts of user cognitive behavior, personality, and attitudes on
the service adoption (Abouelela et al. 2022; Bolduc and Alvarez-Daziano 2010; Ben-Akiva
et al. 2002), but also to get a more realistic choice behavior, as pointed out in Raveau et al.
(2010), Bolduc and Alvarez-Daziano (2010).
The first step in HCM is to estimate the latent constructs of the data (namely attitudes,
travel behavior, and personality) using Exploratory Factor Analysis (EFA). We started the
analysis by performing a scree test (Cattell 1966) to decide on the optimum number of fac-
tors. The test showed two factors as the desired number, and we kept attributes with factor
 Transportation

loading 0.4 or larger, based on the sample size and following (Hair et al. 1998). Varimax
rotation was applied to obtain an orthogonal structure between the different factors, and
the polychoric correlation was used as it suits the ordered nature of the data better than
other correlation methods (Holgado-Tello et al. 2010). After deciding on the estimated fac-
tors for each of the question groups, the corresponding discrete outcome model was first
estimated, and the latent variable model was added afterward. Four HCM models were
estimated using Apollo package (Hess and Palma 2019) under the statistical software R (R
Core Team 2023).

Study setup

Munich is the third largest city in Germany, with a population of around one and a half
million and six million inhabitants in the metropolitan area (Aguilera-García et al. 2022).
The city has a strong transportation infrastructure network reflected in many aspects of
the inhabitants’ daily travel behavior, where 80% of the population owns at least one bike,
served by a 1,200 km long bike lanes network and 28,000 bike parking spaces. Also, the
overall city modal shift reflects the solid public transportation culture, where 33% of the
trips are made by cars, 23% by public transportation, and 44% of daily trips are done by
active mobility, walking and biking.6 The city-shared mobility landscape is vibrant, with
different options for carsharing, bikesharing, shared e-scooters, moped scooters, and e-hail-
ing. Munich City demonstrates an excellent example of a case study for carsharing use city,
with the free-floating carsharing service starting in 2011. In 2019, there were around 2100
shared cars on the city streets. Different operators adopt different pricing schemes, such as
pay per minute, hour, and day, and lately, some operators are calculating trip prices based
on trip length (Aguilera-García et al. 2022).

Analysis results

Summary of sociodemographic and travel behavior characteristics

The survey resulted in 1170 valid and complete responses. Table 2 shows the collected
sample demographic characteristics compared to the city of Munich. The collected data
is skewed compared to the city population in terms of age, education, occupation, and
income, which is a direct result of the sampling strategy targeting young users. In general,
the sociodemographic characteristics of the shared mobility users are different from the
ones of the average population, as discussed in “Literature review” section.
In terms of age, 89% of the sample is younger than 36 years old, compared to 40% of
the average city residents age; also, users are highly educated, with 85% of the sample
having at least a bachelor’s degree compared to 26% of the city’s residents. The number
of students in the sample is over-representative in comparison to the city, as 43% of the
sample respondents are students compared to only 4.5% of the city population. Therefore,
the age and occupation of the respondents are reflected in other aspects, such as income
being lower than the city average and the low number of children in the households. As the
focus target group of this research are users younger than 35 years old, we only considered

6
https://​civit​aS.E.​u/​cities/​munich, last accessed 30/05/2023.
Transportation

Table 2  Summary of sample demographics and travel behavior and comparison with the Munich Census
(2011)
Variable Subgroup n (pct%) User Non-user Munich census

Age 18–24 415 (35%) 175 (30%) 240 (40%) (18–29) 27.2%
25–30 521 (44%) 272 (47%) 249 (42%)
31–35 108 (9.2%) 68 (12%) 40 (7%) (30–39) 16.7%
36–40 46 (3.9%) 29 (5%) 17 (3%)
41+ 81 (6.9%) 34 (6%) 47 (8%) (40+) 51.5%
Gender Female 523 (45%) 241 (42%) 282 (48%) 51.70%
Male 648 (55%) 337 (58%) 311 (52%) 48.30%
Education level Masters & PhD 386 (33%) 219 (38%) 167 (28%) (PhD 2.5%)
Bachelor 657 (56%) 309 (53%) 348 (59%) Bachelor/MS: 22.7%
High school or 128 (11%) 50 (9%) 78 (13%) 66.90%
less
Monthly income 500€ or Less 140 (12%) 40 (6.9%) 100 (17%) Avg: 4220 AC/
500€–2000€ 580 (50%) 259 (45%) 321 (54%) household
2000€–4000€ 259 (22%) 159 (28%) 100 (17%)
4000€ and more 192 (16%) 120 (21%) 72 (12%)
Occupation Full time 405 (34.6%) 258 (45%) 147 (25%) Full/part-time 87.1%
Part-time 165 (14.1%) 81 (14%) 84 (14%)
Self employed 43 (3.7%) 14 (2.4%) 29 (4.9%)
Student 510 (43.6%) 208 (36%) 302 (51%) 4.50%
Other 48 (4.0%) 17 (3%) 31 (5%) 8.40%
Children No 1,019 (87%) 491 (85%) 528 (89%)
Yes 152 (13%) 87 (15%) 65 (11%)
Household size 1 441 (38%) 200 (35%) 241 (41%) 50.30%
2 296 (25%) 174 (30%) 122 (21%) 28.80%
3 and more 434 (37%) 204 (35%) 230 (38%) 20.90%
PT ticket* Yes 859 (73%) 407 (70%) 452 (76%)
No 311 (27%) 171 (30%) 141 (24%)
Own bike or Yes 595 (51%) 335 (58%) 260 (44%)
E-bike No 575 (49%) 243 (42%) 333 (56%)
Car access Yes 451 (39%) 243 (42%) 208 (35%) 44%
No 719 (61%) 335 (58%) 385 (65%) 56%
Driving license** Yes 523 (45%) 343 (59%) 180 (30%) 88.90%
No 647 (55%) 235 (41%) 413 (70%) 11.10%
NTotal = 1170 NUser = 578 NNon−User = 593

*Subscription-based tickets; **Valid in Germany

them in the following analysis, excluding all the other users (N = 1044). When comparing
carsharing users with non-users, using a Perason’s Chi-square test (𝜒2) (Pearson 1900), the
differences were found to be significant in terms of users being males, more educated, with
higher income, compared to the average population, full-time occupation, having access
to a car, and owning a driving license that is valid in Germany. This profile of the carshar-
ing user is similar to other shared mobility services in other locations, such as the United
 Transportation

Bus Car as a driver Car as a passenger E-hailing

100%

61 %

61 %
75%

39 %
37 %
37 %

35 %

35 %
34 %
32 %
50%

29 %

28 %
26 %
24 %

24 %
24 %

23 %

22 %
19 %

18 %
16 %

16 %
15 %

15 %

14 %
13 %

13 %
11 %
25%

9%
9%

9%
8%

8%

8%
7%

6%

6%

4%

2%
1%
1%
0%
Pct (%) of use frquency per mode

Shared bike or
Personal bike Shared e-scooter Suburban train
shared e-bike
100%

73 %

73 %
66 %

57 %
75%
52 %

35 %

50%

29 %
26 %

24 %
23 %

21 %
21 %
20 %

19 %

19 %
19 %
17 %

17 %

16 %
15 %
15 %

15 %
14 %

14 %

13 %

13 %
10 %

10 %

25%

9%
9%
8%

7%

5%
4%

2%
2%

2%

1%
1%

1%
0%

Taxi Tram Underground (metro) Walking

100%

79 %
76 %
63 %

56 %
75%

51 %
50 %

36 %

50%
29 %

27 %
23 %

23 %

22 %
20 %

20 %

18 %
18 %
17 %
16 %

15 %

15 %
14 %
13 %

11 %
11 %
10 %

10 %
25%

9%
8%
7%
6%

6%

6%
3%
3%

2%

2%
1%

1%

1%
1%

0%
4 or more

4 or more

4 or more

4 or more
Less than

Less than

Less than

Less than
times per

times per

times per

times per

times per

times per

times per

times per
times a

times a

times a

times a
month

month

month

month

month

month

month

month
once a

once a

once a

once a
Never

Never

Never

Never
week

week

week

week

week

week

week

week
1–3

1–3

1–3

1–3

1–3

1–3

1–3

1–3
Use frequency
Non-user User

Fig. 3  Urban modes use frequency for users and non-users of carsharing services

States, Canada, Great Britain, and Australia (Howe and Bock 2018; Degele et al. 2018;
Raux et al. 2017; Shaheen and Martin 2015; Kim et al. 2015)
Travel behavior is an important factor that impacts users’ adoption of shared mobility
services (Abouelela et al. 2022); therefore, we asked respondents about the frequency of
their use of twelve modes of transport. The majority of the sample can be described as
active PT users, with at least 40% of the sample using PT more than once a week, which
is reflected in their subscription to PT weekly and monthly tickets. The subscription to
PT services reflects various aspects, such as the users’ loyalty to the service or the high
quality of the PT system. Also, younger respondents more actively using PT than their
older counterparts, who tend to use more private cars, was observed in other locations as
well (Chaisomboon et al. 2020). A considerable percentage of users have access to private
car use, as reflected in their car usage. Active travel is evident in the sample, mainly in
the form of walking and personal bike, and not much use of shared micromobility modes.
We further analyzed the modes used by users versus non-users and also made an assess-
ment by gender; see Fig. 3, Tables 10 and 11. The differences in travel behavior between
the genders are well established, where women generally tend to utilize slower transporta-
tion modes like public transport and walk more frequently than men. They generally travel
shorter distances and have more complex trip arrangements, and are more likely to travel
accompanied by children or dependents, facing more challenges related to physical acces-
sibility, safety, and security (Pourhashem et al. 2022; Xu 2020; Tilley and Houston 2016).
Gender is a decisive factor in shared mobility use, and specifically in the case of carshar-
ing, as discussed in “Literature review” section. Therefore, we considered the travel behav-
iour analysis per gender to further investigate these differences and test their impacts on the
carsharing use.
Figure 3 shows the frequency of using the different urban modes for users and non-
users; to assess the significance of these differences, we performed a chi-square test. From
the twelve compared modes, nine were found to have significant differences, and only three
modes did not have significant differences, namely walking, tram, and the underground
metro. Carsharing users were, on average, more frequent users of all other modes than
Transportation

non-users (of carsharing services), except for bus(es). In terms of gender, differences in
mode frequency were limited and were significant in the case of car use as a passenger and
as a driver, shared bike, and taxi; in particular, males used, on average more bikesharing
systems and were more often car drivers, as compared to their female counterparts.

Familiarity with carsharing and carsharing use

In this section, we explore the respondent familiarity with carsharing services, and the way
they use the service. We asked the users to rank their familiarity with the carsharing service
on a four-point scale ranging from: “I do not know about them” to “Very familiar, I know
almost everything about them.” Most users (65%) knew about the service, and around one-
fifth were very familiar with the service. We asked this question as we believed carshar-
ing use is correlated with user familiarity with them, and we wanted to test the familiarity
impact on the different service use aspects as explained in detail in “Analysis results” sec-
tion. Table 9 shows the summary statistics for the familiarity with carsharing services for
each user and non-users, per gender. Results indicated that users generally had a higher
level of familiarity with the service compared to non-users; 88% of users were familiar
with the service as compared to 43% of non-users. It is important to highlight that the 12%
of the users who were unfamiliar with the service reported that they had used carsharing
mainly as passengers. When assessing by gender, there was no significant difference in
terms of knowledge, except that males were very familiar with the service, as compared to
females.
Table 9 shows the summary statistics of the different aspects of use and familiarity of
carsharing services for the different groups; Chi-square tests were used to test the signifi-
cance of the differences between the different subgroups. The majority of users used the
service as passengers, and they used it mainly less than once per week. The major trip
purposes are leisure, visits, work, and shopping. Users were asked about the modes they
replaced the last carsharing trip with, and the top five modes were the underground, car as a
passenger, suburban train, e-hailing service, and car as a driver. These results show poten-
tial for negative impacts, as carsharing trips replace mainly PT trips which might increase
the vehicle kilometer traveled (VKT) on the roads and, subsequently GHG emissions. We
also asked the users to express their willingness to walk to the nearest carsharing vehicle
locations, for which 75% of the users specified that they would walk up to seven minutes
to the pickup location. We also tested the impact of frequency of use on the willingness
to walk, for which no significant results were found. Users’ willingness to walk was uni-
formly distributed among the different use frequencies. Similar results were observed in
fixed-route commercially organized pooled rides (Abouelela et al. 2022).

Modeling results

In this section, we first present the exploratory factor analysis results, after which we pre-
sent the findings extracted from the four developed hybrid choice models. The aim was to
first extract the latent constructs on both user and service-related aspects to carsharing, to
then incorporate them and assess their impact on carsharing. In particular, the impact of
personality traits and attitudes on knowledge about carsharing, carsharing adoption, and
 Transportation

use, was assessed. The importance of service-related attributes on the choice between car-
sharing operators with different payment schemes was also explored.

Exploratory factor analysis

In this sub-section, the exploratory factor analysis (EFA) results are presented, based on
which the latent constructs have been extracted, notably for user attitudes; the impact of the
extracted factors on carsharing use was then studied. In particular, the factor analysis was
conducted for three question groups relating to carsharing operator-related features (“Car-
sharing operator-related features” section), personality traits (“Personality traits” section),
and travel behavior (“Travel behavior” section).

Carsharing operator‑related features

For the first questions group, we asked respondents to rate how important different aspects
of carsharing services were to them, on a five-point Likert (Likert 1932) scale that ranges
from ( 1 = not important at all, 2 = not important, 3 = neutral, 4 = important, 5 = very
important). Table 10 presents the summary statistics for the ratings of the seven examined
aspects of the carsharing service characteristics. The rating summary shows no significant
difference between gender groups in the evaluation rate; however, a slight difference in
the ranking of the importance of each aspect was observed. Application ease of use was
selected as the most critical aspect, while the availability of EVs in the carsharing fleet
was rated as the least important factor as per the evaluation order for both genders; the lat-
ter was found to be less than neutral for male users with an average evaluation score being
less than 3. The rest of the operator-related features were almost the same for both genders,
with women’s evaluation scores (in terms of importance) generally consistently higher than
males without any statistically significant difference.
When comparing user and non-user groups, interestingly, non-users had, on average
higher evaluation scores for the different aspects, except for the availability of different size
vehicles, which was the second to last least important aspect based on their rating. Also,
application ease of use was the most important service aspect, with a significant difference
in rating compared to the next important aspect, app rating. The differences between users
and non-users were significant and evident in all aspects, except for service availability in
different cities and for app ease of use.
The top part in Table 3 shows the factor analysis results with two main factors repre-
senting the main latent constructs and explaining 46% of the total data variability. Factor
one can be described as the physical offers, and the second factor as the application-related
factors. The results of the EFA for the carsharing operator-related features could possi-
bly reflect on the important dimensions of the service that operators need to focus on to
achieve a high level of satisfaction among users.

Personality traits

Understanding personality traits is essential for understanding human travel behavior; yet
the impact of such traits on travel behavior is still not well comprehended (Jani 2014).
Also, personality might not be a direct influence on travel behavior, but it dictates a certain
pattern of behavior (Revelle 2007), and it is more likely to be associated with different
Transportation

Table 3  Factor analysis model results

I–Carsharing operator-related features Physical offerings Application

App ease of use 0.92

App rating 0.60
Availability in airport 0.71
Availability of different size vehicles 0.62
Service offers bundles 0.56
Availability in other cities 0.53
Availability of electric vehicles 0.51
Model diagnostics
Factor loadings 1.82 1.38
Proportion variance 0.26 0.20
Kaiser–Meyer–Olkin factor adequacy: MSA = 0.80
Cronbach’s alpha = 0.73
II–Personality traits Adventurous Organized

Adventurous 0.82
Being outdoor 0.51
Spontaneous 0.61
Risk taker 0.58
Variety seeking 0.50
Efficient 0.70
Punctual 0.46
Model diagnostics
Factor loadings 1.93 0.76
Proportion variance 0.28 0.11
Kaiser–Meyer–Olkin factor adequacy: MSA = 0.75
Cronbach’s alpha = 0.6
III–Travel behavior Frequent PT user Frequent
micromobility
user

Bikesharing 0.75
Shared E-scooter 0.70
Tram 0.68
Underground 0.85
Suburban Train 0.73
Bus 0.69
Model diagnostics
Factor loadings 2.43 1.10
Proportion variance 0.35 0.16
Kaiser–Meyer–Olkin factor adequacy: MSA = 0.78
Cronbach’s alpha = 0.72
 Transportation

levels of mobility; for example, having an adventurous personality might be associated

with a higher level of mobility (Redmond 2000). The middle part in Table 10 presents the
summary statistics for the answers pertaining to personality traits for different respond-
ent groups (users, non-users, males, and females). In particular, respondents were asked
to specify their agreement with different personality types on a five-point Likert scale
(ranging from “Totally disagree”, “Disagree”, “Neutral”, “Agree”, “Totally agree”). After
conducting Chii-square tests for assessing the statistical significance in personality traits
between different respondent groups (see Table 11 middle part), we found significant dif-
ferences in personality traits between users and non-users than between the different gen-
der groups (i.e., males and females).
Our initial hypothesis for the EFA was that we would estimate five factors representing
the five major personalities, namely risk-taking, loner, ambitious, organized, and lazy. The
middle part in Table 3 presents the estimated EFA results for the personality-related ques-
tions, for which two main factors were extracted, interpreted as “adventurous” and “organ-
ized”. The two factors explain 39% of the data variability. The results of these factors were
further used to estimate the impact of these two types of personalities on carsharing use.

Travel behavior

The final set of questions that were analyzed using EFA techniques focused on the fre-
quency of use of the different available modes. For this question, we hypothesized three
types of users: PT users, private mode users, and finally, shared mobility users. The bottom
part in Table 3 bottom part presents the results of the EFA for the mode use frequency. Two
factors were extracted and found to be significant, one for PT users and the other for shared
micromobility users; the two factors explained 51% of the variance of the data, and the ini-
tial hypothesis was partially correct.

Factors impacting knowledge about carsharing

This model investigates the factors impacting user’s knowledge regarding carsharing. The
answer to the question investigating the knowledge about carsharing was set as the depend-
ent variable, which is ordered in nature. The answers to this question were “I do not know
about them”; “I have heard about them”; “know about them, but not much details”; “Very
familiar, I know almost everything about them”. Ordered HCM model was estimated, and
Fig. 4, and Table 4 show the full path diagram and the estimated model results.
Four variables and two latent variables were significant with positive estimated coeffi-
cients (+𝛽 ), which show that these variables are associated with a higher likelihood regard-
ing higher knowledge about carsharing services: previous use of carsharing, ownership of
a driving license, full-time workers, people who live in small households, adventurous per-
sons, and frequent PT users. The thresholds between the different knowledge levels are
significant, showing that people understand the difference between the different levels.
The latent variable models can be interpreted as follows: for the measurement model
adventurous personality, the positive sign for the estimated coefficient (𝜁 ) for the measure-
ment model part shows that the more the person agrees with the statement, the more likely
is this personality type, and the more likely he is to be an adventurous person. The signs of
the coefficients of the Structure model part (𝛾 ) for males and bike owners show that these
variables increase the probability of being an adventurous person compared to the other
population group. The other latent variable is the PT frequent user, and the measurement
Transportation

Fig. 4  Full path diagram for the ordered HCM for knowledge about carsharing

model positive coefficient (𝜁 ) sign shows that the higher the answer the more frequently
the person uses PT, and the negative sign for the high-income coefficient (𝛾 ) shows that
high-income people are less likely to be frequent PT users. The estimated model partially
answers the first research question (RQ1).

Factors impacting carsharing adoption

This section presents the model results for the model investigating the factors that impact
the adoption of carsharing services, and partially answers RQ1. A binary choice and latent
variable HCM was estimated to investigate the examined factors. For the subject model,
the dependent variable was coded as a binary variable considering responses indicating
that they never used carsharing as zero, with the rest of users being coded as 1.
Figure 5 and Table 5 present the full path diagram and the estimation results for the
hybrid choice model for carsharing adoption. The estimated model shows that people
familiar with carsharing services, with a driving license, who are full-time workers, own-
ers of bikes, with a high-income level, and with a higher education level are more likely to
adopt carsharing services compared to other population groups. These significant varia-
bles are aligned with the hypothesized profile of shared mobility users, who are in general,
wealthier and more educated than the average population. On the other hand, people who
have access to a car, live in a small household, and have a subscription to PT tickets are
less likely to adopt carsharing services. The two latent variables, frequent shared micro-
mobility users (𝜆1) and adventurous personality (𝜆2), were found to be significant predic-
tors impacting the adoption of carsharing. This model shows that users with adventurous
personality have a higher probability of adopting carsharing; such personality was previ-
ously (in previous studies) associated with a preference for higher levels of mobility, being
outdoor, and disliking routine (Gao et al. 2017; Redmond 2000), which might be the utility
provided by carsharing. The other latent variable shows that frequent micromobility users
 Transportation

Table 4  Ordered HCM results for knowledge about carsharing (ordinal variable)
Variable 𝛽 S.E P value

Carsharing use (yes vs. no) 2.09 0.15 0.00

Driving license (yes vs. no) 0.68 0.14 0.00
Occupation (full-time vs. other) 0.34 0.14 0.01
Household size (1–2 vs. 3 and more) 0.21 0.13 0.10
LV1: PT user (𝜆1) 0.20 0.08 0.01
LV2: Adventurous (𝜆2) 0.29 0.08 0.00
Threshold

I do not know about them—I have heard about them − 0.93 0.13 0.00
I have heard about them—I know about them, but not details 0.63 0.12 0.00
I know about them, but not details—very familiar with them 3.73 0.17 0.00
Number of observations = 1044
Latent variable Model

Structure model (Frequent PT user) 𝛾 S.E. P value

Income (2000€ or more vs. less than 2000€) − 0.44 0.08 0.00
Measurement model (Frequent PT user) 𝜁 S.E. P value
Tram 1.68 0.12 0.00
Underground 2.68 0.25 0.00
Suburban train 2.00 0.15 0.00
Bus 1.62 0.13 0.00
Latent variable Model

Structure model (Adventurous personality) 𝛾 S.E. P value

Gender (male vs. female) 0.14 0.07 0.06
Bike or E-Bike ownership (yes vs. no) 0.12 0.07 0.10
Measurement model (Adventurous personality) 𝜁 S.E. P value
Adventurous 2.61 0.24 0.00
Spontaneous 1.37 0.11 0.00
Outdoor 1.13 0.11 0.00
Risk taker 1.43 0.11 0.00
Variety seeker 1.03 0.10 0.00

are more likely to adopt carsharing services in comparison to other population groups. This
behavior was also observed in the adoption of other shared mobility services, such as in the
case of pooled rides (Abouelela et al. 2022).
The lower part of Table 5 shows the structural equation model of the HCM. The estima-
tion of the latent variable model for the personality part shows that the coefficients of the
measurement model part (𝜁 ) is positive, which indicates that the higher the level of agree-
ment with the personality statement questions, the more likely the person to be adventur-
ous. Coefficients of the Structure model (𝛾 ) are positive, showing that each of males and
bike owners (as opposed to females and non-bike owners) are more likely to be adven-
turous. The estimation of the second latent variable model shows that the coefficients of
measurement models (𝜁 ) are positive, indicating that the higher the frequency of using
bike-sharing and/or shared e-scooters, the higher the likelihood to be a frequent shared
Transportation

Fig. 5  Full path diagram for the binary HCM for carsharing adoption

micromobility user. Finally, the (𝛾 ) coefficient for the Structure model part shows that users
who are familiar with carsharing use are more likely to be users of shared micromobility,
and car owners are more likely to use micromobility in comparison to other population
groups, which matches the general profile of shared mobility users. For both latent models,
we did not show the estimation results of the thresholds between the different indicators, as
they have no meaning by themselves and only indicate the order of the thresholds.

Factors impacting the shift to carsharing

This model investigated factors impacting the shift from different modes to carsharing. We
grouped the modes replaced by carsharing into two groups; the first one being the low-
capacity vehicles groups (including cars as a driver, cars as passengers, E-hailing, and
Taxis) and the second group being the PT group (with bus, tram, underground, and subur-
ban trains). These observations amounted to 478 users who shifted from the previous spe-
cific modes, representing 93% of the total number of carsharing users (515 users). The rest
of the observations (37) were removed from the sample used to estimate this model. The
dependent variable of the model was coded as a binary variable with the value of one in
the case of the shift taking place from a low capacity vehicle (cars as a driver, cars as pas-
sengers, E-hailing, and Taxis), the first group, and zero otherwise, similar to the approach
adpoted by Abouelela et al. (2022). Table 6 shows the model estimation results, and Fig. 6
shows the model’s full path diagram.
The estimated model results show that high-income individuals, who are full-time
employed, have access to a car, and are willing to walk less than 5 min to carsharing pick-
up locations, are more likely to shift to carsharing from low-occupancy vehicles as com-
pared to the rest of the population, which are in line with the profile of shared mobility
users. Only one latent variable was significant in this model, namely the frequent PT users.
The negative sign for the latent variable, LV1 (𝜆), showed that PT frequent users are less
likely to shift from low-capacity vehicle trips to carsharing. Similar results were found in
the case of pooled rides, where PT frequent users were less likely to adopt shared mobil-
ity in the form of pooled rides (Abouelela et al. 2022). The latent variable model shows
 Transportation

Table 5  Binary HCM results for carsharing adoption

Variables 𝛽 S.E P value

Intercept −2.40 0.36 0.00

Carsharing familiarity (high vs. low) 2.03 0.17 0.00
Access to car (yes vs. no) − 0.46 0.20 0.01
Driving license (yes vs. no) 1.10 0.17 0.00
Occupation (full-time vs. other) 0.47 0.20 0.01
Education (Ph.D & Masters vs. less than uni degree) 0.60 0.29 0.02
Education (bachelor vs. less than uni degree) 0.49 0.27 0.04
Household size (1–2 vs. 3 and more) − 0.21 0.16 0.10
Bike or e-Bike ownership (yes vs. no) 0.25 0.16 0.06
PT subscription (yes vs. no) − 0.27 0.21 0.10
Income (2000€ or more vs. less than 2000€) 0.32 0.20 0.05
LV1: Frequent micromobility user (𝜆1) 0.28 0.11 0.00
LV2: Adventurous (𝜆2) 0.10 0.09 0.13
𝜌2 = 0.291 𝜌2Adjusted = 0.287
Number of observations = 1044
Latent variable model

Structure model (adventurous personality) 𝛾 S.E. P value

Gender (male vs. female) 0.15 0.07 0.02
Bike or E-Bike ownership (yes vs. no) 0.10 0.07 0.08
Measurement model (adventurous personality) 𝜁 S.E. P value
Adventurous 2.67 0.25 0.00
Spontaneous 1.34 0.11 0.00
Outdoor 1.11 0.10 0.00
Risk taker 1.42 0.11 0.00
Variety seeker 1.02 0.1 0.00
Latent variable model

Structure model (Frequent micromobility user) 𝛾 S.E. P value

Carsharing familiarity (high vs. low) 0.34 0.09 0.00
Bike or e-Bike ownership (yes vs. no) 0.35 0.08 0.00
Measurement model (Frequent micromobility user) 𝜁 S.E. P value
Shared E-scooter 3.76 1.28 0.00
Bike sharing 1.30 0.16 0.00

P values are based on the robust standard errors used to control for heteroscedasticity that might exist

that for the measurement model part, all the coefficients (𝜁 ) are positive, showing that the
higher the use frequency, the more likely it is to be a frequent PT user, which is intuitive.
The Structure model part shows that people who are familiar with carsharing services are
more likely to be frequent PT users, and people who own bikes are more likely to use PT
in comparison with those who do not own bikes. The estimated model answers the the
remaining part of RQ1.
Transportation

Fig. 6  Full path diagram for the binary HCM for shift to carsharing

Table 6  Binary HCM results for the shift from different modes to carsharing
Variable 𝛽 S.E P value

ASC 1.04 0.27 0.00

Willingness to walk to carsharing (less than 5 min 0.32 0.19 0.06
vs. more than 5 min)
Income (2000€ or more vs. less) 0.45 0.24 0.05
Access to car (yes vs. no) 0.66 0.21 0.00
Occupation (Full-time vs. other) 0.46 0.23 0.05
LV1: Frequent PT user (𝜆1) − 0.30 0.12 0.01
𝜌2 = 0.155 𝜌2Adjusted = 0.147
Number of observations = 478
Latent variable Model

Structure model (Frequent PT user) 𝛾 S.E. P value

Bike or E-Bike ownership (yes vs. no) − 0.21 0.11 0.05
Carsharing familiarity (high vs. low) 0.27 0.17 0.09
Measurement model (Frequent PT user) 𝜁 S.E. P value
Tram 2.06 0.21 0.00
Underground 2.57 0.32 0.00
Suburban train 2.26 0.25 0.00
Bus 1.76 0.20 0.00

P values are based on the robust standard errors used to control for heteroscedasticity that might exist

Factors impacting the choice between carsharing operators

This model targeted factors impacting the choice between the different operators with dif-
ferent payment schemes, namely payment per minute or payment per kilometer, which
answers the second research question (RQ2). As shown in Fig. 2, six options were avail-
able; certainly-A and probably-A, indifferent, probably-B, certainly-B, and “None”.
Options certainly-A and probably-A were aggregated to A, the same aggregation was done
 Transportation

Fig. 7  Full path diagram for the multinomial HCM for carsharing operator choice

for options B, the indifferent option was deleted, and option “None” was kept as the third
option, following similar procedures to Abouelela et al. (2021), Fu et al. (2019), Vermeu-
len et al. (2008).
The indifferent options represented 9.3% of the total answers, and the choices of the
remaining aggregated scenarios were distributed as 53.1% for option A, 33.6% for option
B, and 4% for the none option. Our hypotheses for the model-building process were that
males and people who have adventurous personalities might opt for operator B for its pos-
sibility to have cost savings; also, we believe that adventurous users would opt for option B
as they were expected to drive faster for cost saving.
Figure 7 shows the full path diagram and Tables 7 and 8 show the estimated model coef-
ficients and parameters for the HCM of the payment schemes. The interpretation of the
model results considers the “non-trip” option as the reference level for comparison with
other options. The choice experiment tested the significance of four carsharing-related
attributes on the choice between the payment schemes; trip cost, access distance, rating on
the app, and vehicle engine type, electric or not. All the variables were significant except
the access distance. The cost coefficient for option B (pay-per-minute option) was based on
the average cost shown in the experiment, and the coefficient of the vehicle being electric
or not was generic for both options. Interestingly, app rating was the variable with the high-
est absolute coefficient value for this group of variables.
The cost coefficient shows that users value the cost of paying per minute to be cheaper
than paying per km; we believe that this is most likely due to the fact that there is a chance
to pay a lower cost when choosing to pay per minute. Other factors show that app rating
is more effective in the choice of option A, compared to option B. Six user characteristics
were significant, showing that users with high-income levels, familiarity with carsharing
services, valid driving licenses, and who have used carsharing before, were more likely to
adopt carsharing compared to other population groups. On the other hand, people who live
in small size households and who own bikes were less likely to choose car sharing in com-
parison to other groups. Finally, the two latent variables were only significant for option
B, and they indicated that shared micromobility users were more likely to choose option
B, and people who value the importance of the app were more likely to choose option B.
We believe that the main reasons for this are that shared micromobility trips are paid per
Transportation

Table 7  MNL model results for the choice between different carsharing operators
Variable Operator A (per km) Operator B (per min.) None
𝛽 S.E. P value 𝛽 S.E P value 𝛽 S.E. P value

ASC − 4.77 0.20 0.00

Cost (€) − 0.37 0.01 0.00 − 0.32 0.01 0.00
Access distance (m)
Rating on the app. store (⋆) 0.41 0.02 0.00 0.30 0.02 0.00
Electric Vehicle (yes vs. no)* 0.16 0.03 0.00 0.16 0.03 0.00
Income 2000€ or more (vs. less) 0.34 0.13 0.01 0.41 0.13 0.00
Driving license (yes vs. no) 0.57 0.13 0.00 0.43 0.13 0.00
Carsharing familiarity (high vs. 0.64 0.11 0.00 0.38 0.11 0.00
low)
Carsharing use (yes vs. no) 0.79 0.13 0.00 0.83 0.13 0.00
Household size = 1–2 (vs. 3+) − 0.19 0.12 0.10 − 0.26 0.12 0.03
Bike ownership (yes vs. no) − 0.38 0.11 0.01 − 0.46 0.11 0.00
LV1: Frequent micromobility 0.13 0.03 0.00
user (𝜆1)
LV2: Carsharing app feature (𝜆2) 0.14 0.03 0.00
𝜌2 = 0.344 𝜌2Adjusted = 0.343
Number of observations = 9469

P values are based on the robust standard errors used to control for heteroscedasticity that might exist
*Generic coefficient for both options

Table 8  Latent variable model results for the choice between different carsharing operators
Latent variable model

Structure model (Frequent micromobility user) 𝛾 S.E. P value

Gender Male (vs. female) 0.291 0.032 0.000
Measurement model (Frequent micromobility user) 𝜁 S.E. P value
Shared E-scooter 1.973 0.131 0.000
Bike-sharing 1.825 0.093 0.000
Latent variable Model

Structure model (Perceived app importance) 𝛾 S.E. P value

Gender: male (vs. female) − 0.112 0.025 0.000
Income: 2000€ or more (vs. less than 2000€) − 0.056 0.027 0.041
Measurement model (Perceived app importance) 𝜁 S.E. P value
App ease of use 4.052 0.377 0.000
App rating on app store 1.471 0.046 0.000

P values are based on the robust standard errors used to control for heteroscedasticity that might exist
 Transportation

minute of use; besides, people who value the importance of the app in the service users are
more likely to be used to the scheme of paying per minute, which was the original offer for
all the shared vehicle services.
Table 8 shows the latent variable models. The first latent variable model, the importance
of app-rating, can be interpreted as the coefficient (𝜁 ) for the measurement model being
positive, showing that the higher the rating for the importance of app ease of use and the
higher the rating on the app store, the more likely the person is to be in this user group. The
structural part of the model shows that males and high-income individuals are less likely
to be in this group in comparison with the rest of the population. In the second latent vari-
able model, frequent shared micromobility users, the measurement model part coefficients
(𝜁 ) shows that the more frequently shared micromobility used, the more likely to be in this
group. The structural model part shows that male users are more likely to increase the use
of shared micromobility in comparison to female users, which is usually observed in the
case of shared mobility services.
It is important to highlight that our initial hypotheses were not significant and personal-
ity traits did not impact the choice for the payment scheme, and gender indirectly impacted
the choice between payment schemes through the latent variable.

Discussion, limitations, and conclusions

Discussion

In this research, we collected user and carsharing-related data to understand the impact
of psychological factors including personality traits, travel behaviour, and attitudes on the
knowledge about carsharing, its adoption, and use on the one hand, as well as examine
the factors impacting the choice between different carsharing operators.The research was
applied to a case study in Munich, Germany, focusing on young users. The collected data
shows that carsharing users are young, highly educated males with high-income levels,
with full-time jobs, living in small size households, and with a valid driving license, which
is aligned with the general profile of shared mobility services and specifically carsharing
users (Liao et al. 2020; Namazu et al. 2018). Obviously, the characteristics of carsharing
users show the potential for inequitable use problems, wherein population groups, such as
low-income and low-education groups, are not frequent carsharing users, which was evi-
dent in the collected sample, and revealed by the analysis process and the estimated mod-
els. Shared mobility needs a smartphone, digital banking options, and knowledge about the
app use to use the service. Such conditions are not always available and add to the inequi-
table use situation that might result from other conditions, such as service unavailability
within reach and service unaffordability (Abouelela et al. 2024). Digitalization therefore
becomes a concern as it is often highlighted as a key enabler to sustainable development of
cities (Balogun et al. 2020) in general, and to shared mobility in particular (Goehlich et al.
2020). Several strategies could help mitigate this, such as subsidizing the service and offer-
ing an alternative to digital access and digital banking options; however, these solutions do
not always guarantee success. For example, in Chicago, IL, only 0.05% of shared e-scooter
trips were made with non-digital banking options that were provided to help solve the ineq-
uitable use problem for shared e-scooter use (Abouelela et al. 2023). While providing alter-
natives to digital solutions might be plausible in the short-term, addressing concerns of
Transportation

digital literacy and access might be the only viable long-term solution, so that all popula-
tion groups can have access to the service and its digital platform.
The collected data analysis showed that users and non-users have distinguished travel
behavior with significant differences, which indicates the need for further investigation into
how to adjust carsharing service operations to cater to the different travel behaviors and
to attract non-users, if possible. Most of the users (40%) indicated that their last carshar-
ing trip replaced PT (underground, suburban train), showing that there is a potential that
carsharing might increase the VKT, as it replaces large occupancy vehicles (PT). On the
other hand, 35% of users reported carsharing as a replacement for low-occupancy vehicles,
including private cars as passengers or drivers and e-hailing, which may reduce the total
VKT. The latter could have positive impacts such as reducing energy consumption and
resulting CO2 emissions, and required parking spaces (6t-Bureau de recherche and ADEME
2016; Baptista et al. 2014). More information is required, including the access and egress
modes, and the vehicle capacity and occupancy, to better quantify the impacts of carshar-
ing on the VKT; which was not investigated, as it was not the focus of this research.
The responses to the questions regarding familiarity with carsharing services show that
there is a proportional relation between carsharing use and knowledge about the service,
indicating that to increase the use of such services, more marketing and reach-out plans
should be conducted by providers to increase people’s knowledge and awareness regarding
the service, mainly to target non-users.
The EFA was conducted on the three main question groups (service aspect rating, per-
sonality traits, and travel behavior), and each of these groups showed two factors. The first
question group related to the carsharing service’s important aspects showed two factors: (I)
the app-related attributes and (II) physical offers. These estimated factors show the impor-
tance of the app-related attributes, which were not examined in previous research, up to
the best of our knowledge, and which need more investigation to reach the recommended
design by users, as it has a role in impacting service use, as shown in the estimated models.
App-related attributes were significant in the preference of paying per minute, but physical
attributes were not significant in any of the estimated models, confirming the importance
of the app-related aspects of the service.
The second question group is the personality trait group, which showed two distinctive
personality traits, (III) an adventurous personality and (IV) an organized personality. Our
hypothesis was that an adventurous personality would be more likely to use carsharing ser-
vices than other types of personality due to the higher levels of mobility and independence
provided by carsharing, which fits the characteristics of the adventurous personality (Red-
mond 2000). The estimated model showed the significance of the adventurous personality
in adopting carsharing services and the higher level of knowledge regarding the service.
For the last question group, travel behavior, two estimated attitudes were related to
travel behavior; (V) PT frequent user and (VI) shared micromobility user. Both factors
indicate a distinguished travel pattern that shapes the adoption and use of carsharing
services. Shared micromobility users are likelier to adopt the service and prefer to pay
per minute of use, while frequent PT users are less likely to shift from low-capacity
vehicles to carsharing. The impacts of the travel behavior latent construct on the use
of shared mobility use were evident in the case of pooled rides (Abouelela et al. 2022),
showing the importance of accounting for the different travel preferences when plan-
ning new services or even integrating them with current services such as PT, and other
shared services that could increase the potential of multimodality.
Frequent shared micromobility users, in this case, shared e-scooter and bikesharing,
are more likely to adopt other shared mobility services, which highlights the question of
 Transportation

the impacts of shared mobility frequent use on Mobility as a Service (MaaS) platforms
adoption or would the availability of all the shared service within one platform increase
the use of these services, and increase the possibilities of multimodality, which could be
a sustainable outcome. Multimodality is one of the expected positive potential outcome
of MaaS, and subsequently increasing the sustainability of the transport system (Ho and
Tirachini 2024). It is also to be noticed that carsharing service plays a significant role
in MaaS use and utilization, which was observed in the aces of the Augsburg, Germany
MaaS trial, where customers of the Maas bundle utilized their full carsharing allowance
and subsequently increase their carsharing use showing the pivotal role for carsharing in
MaaS use and utilization (Reck et al. 2021). Also, Keller et al. (2018) observed that car-
sharing user have higher intention to use MaaS platforms then the rest of the population.
The estimated models showed that sociodemographics attributes, knowledge about
carsharing, and personal attitudes and personality traits play significant roles in carshar-
ing use. The estimated model showed that the attributes that increase the probability of
carsharing service adoption are: high familiarity with carsharing service, having a valid
driving license, full-time employment, a high education level, high-income level, own-
ing a bike, having an adventurous personality, and being a frequent micromobility user.
The results of this model are in line with the general profile of shared mobility users (Le
Vine and Polak 2019; Martin and Shaheen 2011a; Alemi et al. 2018; Ahmed et al. 2021;
Luo et al. 2019). It is to be noted that the variable with the highest estimated coefficient
is familiarity with carsharing services, followed by the availability of a driving license
and the (high) level of education. It is clear that knowledge about the service is very
important in impacting its adoption, which highlights the role of marketing in service
use. Also, shared mobility users are more likely to use such services in different forms.
On the other hand, users who have access to a car, users with PT subscription-based
tickets, and living in small size households are more likely not to use the service, show-
ing that there is a need to investigate the potential of integrating carsharing services in
the PT subscription to increase the service use.
Again, sociodemographic characteristics and attitudes play a significant factor in the
shift from different modes to carsharing, where high-income people who are full-time
employed, willing to walk for a short period (less than 5 min) and have access to a car
have a higher likelihood to shift from low occupancy vehicles to carsharing, while PT
frequent users are less likely to do so. This model also shows the significance of soci-
odemographics and travel behavior in replacing different modes with carsharing ser-
vices, and it is also in line with the profile of shared mobility users.
When looking at factors impacting the choice between operators with different payment
schemes, trip cost, rating on the app store, and availability of electric vehicles were found to
be quite significant. App rating was the coefficient with the highest reported value, showing
its importance in the choice between different payment schemes. Also, people perceive the
payment per minute as cheaper than the payment per km, which is an interesting result show-
ing the preference of users for the payment scheme per minute (the oldest, more common
scheme for carsharing payment) over the payment per km with all the other factors being
constant. Also, sociodemographics are crucial in choosing between operators, such as high
income, driving license, familiarity, and previous use of carsharing services. On the other
hand, having a bike and living in a small size household reduce the likelihood of carsharing
use. The highest estimated coefficient in this model related to user characteristics is the previ-
ous use of carsharing, showing that people who have experience with the service are more
likely to choose to pay per minute if all other factors are kept constant. Attitudes also played
a significant role, wherein respondents who valued the importance of the app and shared
Transportation

micromobility frequent users are more likely to use the service and choose to pay per minute
of use. These findings highlight the preference for the payment per minute and could be used
by operators to increase their demand by focusing on app development and rating.
The answer to the final research question regarding the knowledge about carsharing ser-
vices emphasized again the importance of sociodemographics and attitudes on the level
of knowledge; in particular, previous use of carsharing, availability of a driving license,
living in small size households, and full-time employees were more likely to have a higher
level of knowledge regarding carsharing service. Service adoption and knowledge about
the service were found to be significant in increasing the probability of each other, show-
ing the need to advertise the service to attract more users and to focus on the other social
groups that do not have enough knowledge regarding the service and subsequently who do
not adopt it. Also, frequent PT users and people with adventurous personalities were more
likely to have a higher knowledge regarding the service. Two highlights from these findings
are that frequent PT user knowledge about the service should be coupled with encouraging
carsharing use as a first-last mile solution that could increase multimodality.

Study limitations and future research needs

This research tries to update the current knowledge regarding carsharing services, using
a mix of revealed answer questions and a stated preference experiment. The study comes
with limitations, which we believe do not impact the overall research integrity. The
main objectives of appraising the limitations are to have a transparent outcome and to
help similar studies avoid or consider them in the future. The collected sample was bal-
anced in terms of users versus non-users of carsharing services and in gender, but it was
unbalanced for other sociodemographic characteristics, such as income level and educa-
tion level. On the other hand, shared mobility users are likely to be young and highly
educated compared to the average population, which makes the sample acceptable for
the purpose of the study, and the sample was not representative of the city’s population.
The findings should not be directly interpolated or carried out on other social groups.
Different attitudes were examined, along with their impacts on the different aspects of
carsharing use, but it is important to be noted that attitude and personality traits are
hard to quantify and measure. They are essential to understand user preferences for the
different aspects of shared mobility use, and they might be more significant and influen-
tial in deciding travel behavior in general and shared mobility use.
The used stated preference experience examined only a number of attitudes, travel
cost, app rating, electrification of the vehicle, and access distance to the nearest vehi-
cle; other attributes could have been used as well, but this was done on purpose, not
to overload the respondents with information that might distract their attention, and
to have a simpler experience. The stated preference experiment assumed that the pay-
ment by KM is a fixed cost, although this can slightly change in reality, such as in the
case of congestion; users could alternate from the original route, the shortest path,
causing extra travel distance that would increase the trip cost. The variation of the
travel cost ( ±25%) around the average trip value would cover this possibility. The sur-
vey was deployed online, which can create a response bias, as groups with no access
to the internet and older populations might not be represented in the sample, but the
well established shared mobility user profile shows that they are mostly young and
highly educated individuals with access to the internet. The hybrid choice models are
not the only way to implement attitudes into discrete models, but we believe that in
 Transportation

this research, they fit the required methodology to answer the main research questions.
The personality traits that were estimated via EFA were what the people report, their
self-perception on their own personality, but might not be how they are if they had
done real psychometric tests. Finally, the assessment of the impact of modal shift (to
carsharing) on VKT was not conclusive (see “Discussion” section), as in most cases
carsharing trips replaced PT (likely increasing VKT), at the same time they also often
replaced small occupancy vehicles such as cars (possibly reducing thereby VKT). To
further investigate this and better quantify the impact, more information would be
needed regarding the trips replaced, such as trip distance, vehicle occupancy, and the
modes used to access and egress the carsharing services. To project the findings on
a larger scale, additional travel behavior data would be essential, so that the modal
shift analysis does not only rely on the last trip made, but rather go beyond it to take
into account a longer time frame which would encompass the frequency at which such
modal shift would occur. As the above was not part of this study, a further in-depth
exploration for the VKT analysis is recommended for future research.
As currently carsharing only accounts for a small portion of the total modal share
compared to private cars, the magnitude of its impacts is limited (Migliore et al. 2020).
Future research could also focus on how extending the service coverage areas, fleet
size, and ideally electrifying the fleet could help cities reap the optimum benefits of
carsharing (Migliore et al. 2020; Harris et al. 2021; Ye et al. 2021).
It is important to highlight that the survey data was collected during the last waves
of the COVID-19 pandemic, and it should be noted that the pandemic conditions inevi-
tably impacted carsharing use and safety perception on different levels. Previous stud-
ies on the pandemic impact on carsharing use has been inconclusive, for instance, in
Madrid, Spain, carsharing has been perceived by some users as a means to avoid pub-
lic transport (and therefore as a safer mode), while for others less so, as they replaced
it with walking and biking (Alonso-Almeida 2022). A study in Poland showed other
findings, in which the pandemic was not a challenge for carsharing users, as it did not
hinder their overall use (Gorzelańczyk et al. 2022).

Conclusions

This research investigated the impacts of personality traits and attitudes on the different
aspects of carsharing use: adoption, the shift from other modes, the choice between differ-
ent operators, and finally, the knowledge about the carsharing services. A large sample (N
= 1044) of young user data was used in the analysis collected from Munich, Germany. The
results continue to highlight the importance of the user sociodemographic characteristics in
impacting service use and raise questions regarding inequitable service use and adoption.
The findings of the estimated econometric models also show the significance of personal-
ity traits, travel behavior, and digital service aspects (such as app ease of use and rating on
the app store) on carsharing use. These findings also stress the importance of designing
user-friendly apps and maintaining good ratings, which can attract more users. Findings
also showed that frequent shared mobility users adopt shared mobility in different forms of
the service, showing the potential of MaaS in increasing shared mobility use and increas-
ing the potential of multimodality. Finally, the estimated models could be used as a part of
broader travel demand models that could estimate the adoption of carsharing and which
might be used to quantify the share of the operators based on their payment methods.
Transportation

Appendix A: Additional analysis

See Tables 9, 10 and 11.

Table 9  Usage and familiarity with carsharing service summary statistics

Male (%) Female (%) Non-user (%) User (%)

Familiarity with carsharing services

I do not know about them 13 13 24 2

I have heard about them 20 23 33 10
I know about them, but not much details 49 50 39 60
Very familiar, I know almost everything about them 18 13 4 28
Willingness to walk

Less than 2 min 12 13 14 11

2–4 min 29 22 22 29
5–7 min 34 35 34 36
8–10 min 21 15 20 17
More than 10 min 10 7 10 8

Type of use

Yes, as a driver 16 9
Yes, as a passenger 24 31
Yes, sometimes as a passenger, and sometimes as a 11 8
driver
Never 49 53
Frequency of use

Never 49 53
Less than once a month 37 34
1–3 times per month 12 11
1–3 times a week 1 1
4 or more times per week 1 0

Replaced Mode (top 5 modes representing 75% of the users)

Underground (U-Bahn) 24 26
Suburban train (S-Bahn) 15 14
Car as a driver 14 11
E-hailing (Uber, and similar) 11 9
Car as a passenger 11 12
Trip purpose (top 4 purposes representing 80% of users who used Carsharing

Leisure (Restaurants, bars, parties) 40 41

Visiting someone 19 18
Work 13 10
Shopping 7 11
 Transportation

Table 10  Attitudinal questions summary statistics

User (mean ± sd) None-user (mean Female (mean Male (mean
± sd) ± sd) ± sd)

Service rating

App ease of use 4.20 ± 0.94 4.21 ± 0.95 4.25 ± 0.97 4.17 ± 0.91
App rating 3.54 ± 1.03 3.74 ± 1.05 3.68 ± 1.06 3.61 ± 1.03
Service offers bundles 3.65 ± 1.19 3.81 ± 1.03 3.81 ± 1.08 3.67 ± 1.15
Availability in other cities 3.52 ± 1.18 3.7 ± 1.16 3.63 ± 1.19 3.60 ± 1.16
Availability in Airport 3.60 ± 1.29 3.71 ± 1.17 3.82 ± 1.21 3.52 ± 1.23
Availability of different size 3.50 ± 1.20 3.45 ± 1.12 3.60 ± 1.15 3.37 ± 1.16
vehicles
Availability of EV 3.02 ± 1.16 3.23 ± 1.15 3.30 ± 1.13 2.98 ± 1.16
Personality traits
Adventurous 0.83 ± 0.85 0.61 ± 1.01 0.66 ± 0.97 0.77 ± 0.92
Anxious − 0.05 ± 1.04 0.05 ± 1.08 0.2 ± 1.03 − 0.16 ± 1.06
Being in Charge 0.72 ± 0.83 0.57 ± 0.89 0.68 ± 0.86 0.62 ± 0.88
Being outdoor 1.06 ± 0.83 0.84 ± 0.98 0.99 ± 0.89 0.91 ± 0.92
Calm 0.68 ± 0.98 0.71 ± 0.95 0.54 ± 0.98 0.81 ± 0.94
Creative 0.74 ± 0.95 0.74 ± 0.93 0.86 ± 0.88 0.64 ± 0.97
Efficient 0.88 ± 0.85 0.81 ± 0.83 0.91 ± 0.83 0.80 ± 0.85
Independent 1.10 ± 0.75 0.99 ± 0.83 1.03 ± 0.79 1.06 ± 0.8
Lazy − 0.13 ± 1.10 − 0.12 ± 1.07 − 0.17 ± 1.04 − 0.09 ± 1.12
Like to be alone 0.15 ± 0.99 0.16 ± 1.11 0.11 ± 1.02 0.2 ± 1.09
Optimistic 0.91 ± 0.85 0.72 ± 0.95 0.76 ± 0.89 0.86 ± 0.92
Participating 0.92 ± 0.72 0.79 ± 0.81 0.85 ± 0.77 0.86 ± 0.77
Punctual 0.66 ± 1.14 0.78 ± 1.05 0.7 ± 1.09 0.74 ± 1.10
Risk taker 0.14 ± 1.02 0.05 ± 1.08 − 0.02 ± 1.02 0.19 ± 1.07
Routines 0.47 ± 0.97 0.45 ± 1.00 0.50 ± 0.98 0.43 ± 0.99
Spontaneous 0.65 ± 0.96 0.54 ± 0.92 0.57 ± 0.94 0.61 ± 0.94
Stay close to home − 0.09 ± 1.04 0.25 ± 1.05 0.11 ± 1.06 0.06 ± 1.06
Variety Seeker 0.78 ± 0.79 0.73 ± 0.86 0.72 ± 0.8 0.78 ± 0.85

Mode use frequency

Bus 2.49 ± 1.28 2.71 ± 1.27 2.6 ± 1.27 2.6 ± 1.28

Car as driver 1.47 ± 1.44 0.9 ± 1.35 0.98 ± 1.31 1.34 ± 1.49
Car as passenger 1.70 ± 1.05 1.42 ± 1.19 1.71 ± 1.17 1.44 ± 1.08
Personal Bike 1.67 ± 1.54 1.19 ± 1.48 1.3 ± 1.51 1.53 ± 1.54
Suburban train 2.23 ± 1.26 2.2 ± 1.37 2.17 ± 1.32 2.25 ± 1.31
Shared E-Scooter 0.7 ± 0.97 0.42 ± 0.81 0.5 ± 0.87 0.61 ± 0.93
Bike sharing 0.54 ± 0.89 0.43 ± 0.84 0.39 ± 0.8 0.56 ± 0.91
Taxi 0.69 ± 0.87 0.5 ± 0.77 0.68 ± 0.9 0.52 ± 0.75
Tram 2.16 ± 1.33 2.11 ± 1.4 2.13 ± 1.38 2.14 ± 1.35
underground 2.97 ± 1.29 3 ± 1.37 3.01 ± 1.34 2.97 ± 1.33
Walking 3.63 ± 0.82 3.63 ± 0.78 3.62 ± 0.82 3.64 ± 0.78
E-hailing 0.94 ± 0.94 0.54 ± 0.81 0.81 ± 0.94 0.68 ± 0.85
Transportation

Table 10  (continued)
Personality question levels were coded as Totally disagree = − 2, Disagree = − 1,
Neutral = 0, Agree = 1, Totally agree = 2
Mode use frequency levels were coded as Never = 0, Less than once a month = 1,
1–3 times per month = 2, 1–3 times a week = 3, 4 or more times per week = 4

Table 11  𝜒 2 test results for personality traits

Users versus None-user Male versus Female

𝜒2 ‘P value’ 𝜒2 ‘P value’

Service aspect rating

Service availability in other cities 6.36 0.17 3.60 0.46
Offers bundles 15.40 0.00* 4.81 0.31
App rating 17.70 0.00* 6.62 0.16
Different size vehicles availability 13.4 0.01* 10.46 0.03*
EV availability 14.80 0.01* 22.00 0.00*
Airport availability 16.40 0.00* 25.73 0.00*
App ease of use 1.68 0.80 11.77 0.02*

Personality

Efficient 6.20 0.18 5.54 0.24

Independent 6.72 0.15 5.57 0.23
Routines 2.57 0.633 1.94 0.75
Punctual 7.38 0.11 3.35 0.50
Variety Seeking 4.62 0.32 4.70 0.32
Lazy 2.04 0.72 5.01 0.29
Stay close to home 32.6 0.00* 3.06 0.55
Being outdoor 20.00 0.00* 3.26 0.52
Spontaneous 8.77 0.06* 3.72 0.45
Being in charge 12.6 0.01* 4.10 0.39
Participative 10.10 0.03* 5.46 0.24
being alone 12.90 0.01* 6.02 0.20
Optimistic 12.20 0.01* 8.29 0.08
Adventurous 20.90 0.00* 8.29 0.08
Creative 1.48 0.83 15.00 0.00*
Calm 6.89 0.14 21.90 0.00*
Risk taker 4.60 0.33 22.70 0.00*
Anxious 5.52 0.23 31.60 0.00*
Mode use frequency
Tram 3.28 0.51 3.37 0.50
Walking 6.23 0.18 6.2 0.19
Underground (metro) 7.10 0.13 1.22 0.88
Bus 9.48 0.05* 4.98 0.29
Suburban train 12.00 0.02* 1.5 0.83
Shared-e-Scooter 29.90 0.00* 6.15 0.19
Personal Bike 31.40 0.00* 7.38 0.12
 Transportation

Table 11  (continued)
Users versus None-user Male versus Female

𝜒2 ‘P value’ 𝜒2 ‘P value’

E-hailing 64.00 0.00* 7.41 0.12

Shard Bike 9.97 0.04* 12.7 0.01*
Taxi 18.90 0.00* 14.5 0.01*
Car as a passenger 43.10 0.00* 16.2 0.00*
Car as as driver 71.50 0.00* 18.2 0.00*

*Significant difference with minimum 95% significance level

Acknowledgements The authors would like to thank Dr.-Ing. Benjamin Büttner, the head of the EIT Urban
Mobility “Doctoral Training Network”, and the DTN for their support. This study was partially funded by
European Union’s Horizon Europe research and innovation program under grant agreement No 101076963
[project PHOEBE (Predictive Approaches for Safer Urban Environment)].

Author contributions M.A., C.A., and C.A conceived and planned the experiments. M.A. carried out the
experiments. M.A. and C.A. planned and carried out the data analysis. M.A., C.A., and C.A. contributed to
models estimation. M.A., C.A., and C.A. contributed to the interpretation of the results. All authors took the
lead in writing the manuscript. All authors provided critical feedback and helped shape the research, analy-
sis, and manuscript.

Funding Open Access funding enabled and organized by Projekt DEAL.

Declarations
Conflict of interest The authors declare no conflict of interest.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com-
mons licence, and indicate if changes were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

References
6t-Bureau de recherche, & ADEME. Enquête Nationale sur l’Autopartage—Edition 2016 Analyse des
enquêtes. Technical Report ADEME (2016)
Abouelela, M., Al Haddad, C., Antoniou, C.: Are young users willing to shift from carsharing to scooter-
sharing? Transp. Res. Part D: Transp. Environ. 95, 102821 (2021)
Abouelela, M., Tirachini, A., Chaniotakis, E., Antoniou, C.: Characterizing the adoption and frequency of
use of a pooled rides service. Transp. Res. Part C: Emerg. Technol. 138, 103632 (2022)
Abouelela, M., Chaniotakis, E., Antoniou, C.: Understanding the landscape of shared-e-scooters in North
America; spatiotemporal analysis and policy insights. Transp. Res. Part A: Policy Pract. 169, 103602
(2023). https://​doi.​org/​10.​1016/j.​tra.​2023.​103602
Abouelela, M., Durán-Rodas, D., Antoniou, C.: Do we all need shared E-scooters? An accessibility-
centered spatial equity evaluation approach. Transp. Res. Part A: Policy Pract. 181, 103985 (2024)
Acheampong, R., Siiba, A.: Modelling the determinants of car-sharing adoption intentions among young
adults: the role of attitude, perceived benefits, travel expectations and socio-demographic factors.
Transportation 47, 2557–2580 (2020). https://​doi.​org/​10.​1007/​s11116-​019-​10029-3
Transportation

Aguilera-García, Á., Gomez, J., Antoniou, C., Vassallo, J.M.: Behavioral factors impacting adoption and
frequency of use of carsharing: a tale of two European cities. Transp. Policy 123, 55–72 (2022).
https://​doi.​org/​10.​1016/j.​tranp​ol.​2022.​04.​007
Ahmed, S., Xu, M., Tsz Ching, H.: From the users’ and the operators’ perceptions: the potential of carshar-
ing in Hong Kong. In: ICIT 2021: 2021 The 9th International Conference on Information Technol-
ogy: IoT and Smart City, pp. 545–553 (2021). https://​doi.​org/​10.​1145/​35125​76.​35126​69
Akyelken, N., Banister, D., Givoni, M.: The sustainability of shared mobility in London: the dilemma for
governance. Sustainability 10, 420 (2018)
Al Haddad, C., Chaniotakis, E., Straubinger, A., Plötner, K., Antoniou, C.: Factors affecting the adoption
and use of urban air mobility. Transp. Res. Part A: Policy Pract. 132, 696–712 (2020)
Alemi, F., Circella, G., Handy, S., Mokhtarian, P.: What influences travelers to use Uber? Exploring the fac-
tors affecting the adoption of on-demand ride services in California. Travel Behav. Soc. 13, 88–104
(2018). https://​doi.​org/​10.​1016/j.​tbs.​2018.​06.​002
Alonso-Almeida, M.D.M.: To use or not use car sharing mobility in the ongoing Covid-19 pandemic? Iden-
tifying sharing mobility behaviour in times of crisis. Int. J. Environ. Res. Public Health 19, 3127
(2022)
Alonso-Mora, J., Samaranayake, S., Wallar, A., Frazzoli, E., Rus, D.: On-demand high-capacity ride-shar-
ing via dynamic trip-vehicle assignment. Proc. Natl. Acad. Sci. 114, 462–467 (2017)
Amirnazmiafshar, E., Diana, M.: A review of the socio-demographic characteristics affecting the demand
for different car-sharing operational schemes. Transp. Res. Interdiscip. Perspect. 14, 100616 (2022)
Arteaga-Sánchez, R., Belda-Ruiz, M., Ros-Galvez, A., Rosa-Garcia, A.: Why continue sharing: determi-
nants of behavior in ridesharing services. Int. J. Mark. Res. 62, 725–742 (2020)
Balac, M., Ciari, F., Axhausen, K.W.: Carsharing demand estimation: Zurich, Switzerland, area case study.
Transp. Res. Rec. 2563, 10–18 (2015)
Balogun, A., Marks, D., Sharma, R., Shekhar, H., Balmes, C., Maheng, D., Arshad, A., Salehi, P.: Assessing
the potentials of digitalization as a tool for climate change adaptation and sustainable development in
urban centres. Sustain. Cities Soc. 53, 101888 (2020)
Baptista, P., Melo, S., Rolim, C.: Energy, environmental and mobility impacts of car-sharing systems.
Empirical results from Lisbon, Portugal. Procedia Soc. Behav. Sci. 111, 28–37 (2014)
Ben-Akiva, M., McFadden, D., Gärling, T., Gopinath, D., Walker, J., Bolduc, D., Börsch-Supan, A.,
Delquié, P., Larichev, O., Morikawa, T., et al.: Extended framework for modeling choice behavior.
Mark. Lett. 10, 187–203 (1999)
Ben-Akiva, M., Walker, J., Bernardino, A.T., Gopinath, D.A., Morikawa, T., Polydoropoulou, A.: Integra-
tion of choice and latent variable models. Perpetual Motion Travel Behav. Res. Oppor. Appl. Chall.
2002, 431–470 (2002)
Bhagat-Conway, M.W., Mirtich, L., Salon, D., Harness, N., Consalvo, A., Hong, S.: Subjective variables in
travel behavior models: a critical review and standardized transport attitude measurement protocol
(STAMP). Transportation 51, 155–191 (2024)
Bocken, N., Jonca, A., Södergren, K., Palm, J.: Emergence of carsharing business models and sustainability
impacts in Swedish cities. Sustainability (2020). https://​doi.​org/​10.​3390/​su120​41594
Bohannon, R.W., Andrews, A.W.: Normal walking speed: a descriptive meta-analysis. Physiotherapy 97,
182–189 (2011)
Bolduc, D., Alvarez-Daziano, R.: On estimation of hybrid choice models. In: Choice Modelling: The State-
of-the-Art and the State-of-Practice: Proceedings from the Inaugural International Choice Modelling
Conference. Emerald Group Publishing Limited, pp. 259–287 (2010)
Burghard, U., Dütschke, E.: Who wants shared mobility? Lessons from early adopters and mainstream
drivers on electric carsharing in Germany. Transp. Res. Part D: Transp. Environ. 71, 96–109 (2019).
https://​doi.​org/​10.​1016/j.​trd.​2018.​11.​011. (The roles of users in low-carbon transport innovations:
Electrified, automated, and shared mobility)
Burghard, U., Scherrer, A.: Sharing vehicles or sharing rides–psychological factors influencing the accept-
ance of carsharing and ridepooling in Germany. Energy Policy (2022). https://​doi.​org/​10.​1016/j.​
enpol.​2022.​112874
Buschmann, S., Chen, M.-F., Hauer, G.: An integrated model of the theory of reasoned action and technol-
ogy acceptance model to predict the consumers’ intentions to adopt electric carsharing in Taiwan.
(2020). https://​doi.​org/​10.​1007/​978-3-​662-​60806-7_9
Cattell, R.B.: The scree test for the number of factors. Multivar. Behav. Res. 1, 245–276 (1966)
Chaisomboon, M., Jomnonkwao, S., Ratanavaraha, V.: Elderly users’ satisfaction with public transport in
Thailand using different importance performance analysis approaches. Sustainability 12, 9066 (2020)
Chen, X., Cheng, J., Ye, J., Jin, Y., Li, X., Zhang, F.: Locating station of one-way carsharing based on spa-
tial demand characteristics. J. Adv. Transp. 2018, 5493632 (2018)
 Transportation

Chicco, A., Diana, M.: Air emissions impacts of modal diversion patterns induced by one-way car sharing:
a case study from the city of Turin. Transp. Res. Part D: Transp. Environ. (2021). https://​doi.​org/​10.​
1016/j.​trd.​2020.​102685
Clewlow, R.: Carsharing and sustainable travel behavior: results from the San Francisco Bay Area.
Transp. Policy 51, 158–164 (2016). https://​doi.​org/​10.​1016/j.​tranp​ol.​2016.​01.​013
Czarnetzki, F., Siek, F.: Decentralized mobility hubs in urban residential neighborhoods improve the
contribution of carsharing to sustainable mobility: findings from a quasi-experimental study.
Transportation (2022). https://​doi.​org/​10.​1007/​s11116-​022-​10305-9
De Lorimier, A., El-Geneidy, A.M.: Understanding the factors affecting vehicle usage and availability in
carsharing networks: a case study of Communauto carsharing system from Montréal. Canada. Int.
J. Sustain. Transp. 7, 35–51 (2013)
Degele, J., Gorr, A., Haas, K., Kormann, D., Krauss, S., Lipinski, P., Tenbih, M., Koppenhoefer, C.,
Fauser, J., Hertweck, D.: Identifying e-scooter sharing customer segments using clustering. In:
2018 IEEE International Conference on Engineering, Technology and Innovation (ICE/ITMC).
IEEE, pp. 1–8 (2018)
Diana, M., Chicco, A.: The spatial reconfiguration of parking demand due to car sharing diffusion: a
simulated scenario for the cities of Milan and Turin (Italy). J. Transp. Geogr. (2022). https://​doi.​
org/​10.​1016/j.​jtran​geo.​2021.​103276
Dias, F.F., Lavieri, P.S., Garikapati, V.M., Astroza, S., Pendyala, R.M., Bhat, C.R.: A behavioral choice
model of the use of car-sharing and ride-sourcing services. Transportation 44, 1307–1323 (2017)
Duncan, M.: The cost saving potential of carsharing in a US context. Transportation 38, 363–382 (2011).
https://​doi.​org/​10.​1007/​s11116-​010-​9304-y
Efthymiou, D., Antoniou, C.: Modeling the propensity to join carsharing using hybrid choice models
and mixed survey data. Transp. Policy 51, 143–149 (2016). https://​doi.​org/​10.​1016/j.​tranp​ol.​2016.​
07.​001
Efthymiou, D., Antoniou, C., Waddell, P.: Factors affecting the adoption of vehicle sharing systems by
young drivers. Transp. Policy 29, 64–73 (2013). https://​doi.​org/​10.​1016/j.​tranp​ol.​2013.​04.​009
Ejelöv, E., Nilsson, A.: Individual factors influencing acceptability for environmental policies: a review
and research agenda. Sustainability 12, 2404 (2020)
Engel-Yan, J., Passmore, D.: Carsharing and car ownership at the building scale. J. Am. Plann. Assoc.
79, 82–91 (2013). https://​doi.​org/​10.​1080/​01944​363.​2013.​790588
Feng, X., Sun, H., Wu, J., Lv, Y.: Understanding the factors associated with one-way and round-trip
carsharing usage based on a hybrid operation carsharing system: a case study in Beijing. Travel
Behav. Soc. 30, 74–91 (2023)
Fleury, S., Tom, A., Jamet, E., Colas-Maheux, E.: What drives corporate carsharing acceptance? A
French case study. Transp. Res. F: Traffic Psychol. Behav. 45, 218–227 (2017). https://​doi.​org/​10.​
1016/j.​trf.​2016.​12.​004
Fu, M., Rothfeld, R., Antoniou, C.: Exploring preferences for transportation modes in an urban air
mobility environment: Munich case study. Transp. Res. Rec. 2673, 427–442 (2019)
Furnham, A., Saipe, J.: Personality correlates of convicted drivers. Personal. Individ. Differ. 14, 329–
336 (1993)
Gao, Y., Rasouli, S., Timmermans, H., Wang, Y.: Understanding the relationship between travel satis-
faction and subjective well-being considering the role of personality traits: a structural equation
model. Transport. Res. F: Traffic Psychol. Behav. 49, 110–123 (2017)
Gilibert, M., Ribas, I.: Synergies between app-based car-related shared mobility services for the devel-
opment of more profitable business models. J. Ind. Eng. Manag. 12, 405–420 (2019)
Goehlich, V., Fournier, G., Richter, A.: What can we learn from digitalisation and servitisation to shape
a new mobility paradigm? Int. J. Bus. Glob. 24, 296–306 (2020)
Golalikhani, M., Oliveira, B.B., Carravilla, M.A., Oliveira, J.F., Pisinger, D.: Understanding carsharing:
a review of managerial practices towards relevant research insights. Res. Transp. Bus. Manag. 41,
100653 (2021)
Gorzelańczyk, P., Kalina, T., Jurkovič, M.: Impact of the Covid-19 pandemic on car-sharing in Poland.
Komunikácie 24, A172 (2022)
Gössling, S.: Integrating e-scooters in urban transportation: problems, policies, and the prospect of sys-
tem change. Transp. Res. Part D: Transp. Environ. 79, 102230 (2020)
Hair, J., Anderson, R., Tatham, R., Black, W.: Multivariate data analysis. In: 5th International Ed (1998)
Harris, S., Mata, E., Plepys, A., Katzeff, C.: Sharing is daring, but is it sustainable? An assessment of
sharing cars, electric tools and offices in Sweden. Resour. Conserv. Recycl. (2021). https://​doi.​org/​
10.​1016/j.​resco​nrec.​2021.​105583
Transportation

Hartl, B., Hofmann, E.: The social dilemma of car sharing-the impact of power and the role of trust in
community car sharing. Int. J. Sustain. Transp. 16, 526–540 (2022)
Hess, S., Palma, D.: Apollo: a flexible, powerful and customisable freeware package for choice model
estimation and application. J. Choice Model. 32, 100170 (2019)
Hjorteset, M., Böcker, L.: Car sharing in Norwegian urban areas: examining interest, intention and the deci-
sion to enrol. Transp. Res. Part D: Transp. Environ. (2020). https://​doi.​org/​10.​1016/j.​trd.​2020.​102322
Ho, C.Q., Tirachini, A.: Mobility-as-a-service and the role of multimodality in the sustainability of urban
mobility in developing and developed countries. Transp. Policy 145, 161–176 (2024)
Holgado-Tello, F.P., Chacón-Moscoso, S., Barbero-García, I., Vila-Abad, E.: Polychoric versus Pearson
correlations in exploratory and confirmatory factor analysis of ordinal variables. Qual. Quant. 44,
153–166 (2010)
Howe, E., Bock, B.: Global scootersharing market report 2018 (2018)
Hu, S., Chen, P., Lin, H., Xie, C., Chen, X.: Promoting carsharing attractiveness and efficiency: an explora-
tory analysis. Transp. Res. Part D: Transp. Environ. 65, 229–243 (2018a)
Hu, S., Lin, H., Xie, K., Chen, X., Shi, H.: Modeling users’ vehicles selection behavior in the urban carshar-
ing program. In: 2018 21st International Conference on Intelligent Transportation Systems (ITSC).
IEEE, pp. 1546–1551 (2018b)
Jain, T., Johnson, M., Rose, G.: Exploring the process of travel behaviour change and mobility trajectories
associated with car share adoption. Travel Behav. Soc. 18, 117–131 (2020). https://​doi.​org/​10.​1016/j.​
tbs.​2019.​10.​006
Jain, T., Rose, G., Johnson, M.: Don’t you want the dream?: Psycho-social determinants of car share adop-
tion. Transp. Res. F: Traffic Psychol. Behav. 78, 226–245 (2021). https://​doi.​org/​10.​1016/j.​trf.​2021.​
02.​008
Jain, T., Rose, G., Johnson, M.: Changes in private car ownership associated with car sharing: gauging dif-
ferences by residential location and car share typology. Transportation 49, 503–527 (2022). https://​
doi.​org/​10.​1007/​s11116-​021-​10184-6
Janasz, T., Schneidewind, U.: The future of automobility. In: Shaping the Digital Enterprise, pp. 253–285.
Springer (2017)
Jani, D.: Relating travel personality to big five factors of personality. Tour. Int. Interdiscip. J. 62, 347–359
(2014)
Jochem, P., Frankenhauser, D., Ewald, L., Ensslen, A., Fromm, H.: Does free-floating carsharing reduce
private vehicle ownership? The case of SHARE NOW in European cities. Transp. Res. Part A: Policy
Pract. 141, 373–395 (2020). https://​doi.​org/​10.​1016/j.​tra.​2020.​09.​016
Jorge, D., Correia, G.: Carsharing systems demand estimation and defined operations: a literature review.
Eur. J. Transp. Infrastruct. Res. 13, 201 (2013)
Jung, J., Koo, Y.: Analyzing the effects of car sharing services on the reduction of greenhouse gas (GHG)
emissions. Sustainability (2018). https://​doi.​org/​10.​3390/​su100​20539
Keller, E., Aguilar, A., Hanss, D.: Car sharers’ interest in integrated multimodal mobility platforms: a diffu-
sion of innovations perspective. Sustainability 10, 4689 (2018)
Kent, J.: Carsharing as active transport: What are the potential health benefits? J. Transp. Health 1, 54–62
(2014). https://​doi.​org/​10.​1016/j.​jth.​2013.​07.​003
Khan, M., Machemehl, R.: The impact of land-use variables on free-floating carsharing vehicle rental choice
and parking duration. In: Seeing Cities Through Big Data. Springer, pp. 331–347 (2017)
Kim, D., Ko, J., Park, Y.: Factors affecting electric vehicle sharing program participants’ attitudes about car
ownership and program participation. Transp. Res. Part D: Transp. Environ. 36, 96–106 (2015)
Kim, D., Shin, H., Im, H., Park, J.: Factors influencing travel behaviors in bikesharing. In: Transportation
Research Board 91st Annual Meeting, pp. 1–14 (2012)
Kireeva, N., Zavyalov, D., Saginova, O., Zavyalova, N.: Car sharing market development in Russia. Transp.
Res. Procedia 54, 123–128 (2021)
Ko, J., Ki, H., Lee, S.: Factors affecting carsharing program participants’ car ownership changes. Transp.
Lett. 11, 208–218 (2019). https://​doi.​org/​10.​1080/​19427​867.​2017.​13298​91
Kolleck, A.: Does car-sharing reduce car ownership? Empirical evidence from Germany. Sustainability
(2021). https://​doi.​org/​10.​3390/​su131​37384
Kroesen, M., Chorus, C.: A new perspective on the role of attitudes in explaining travel behavior: a psycho-
logical network model. Transp. Res. Part A: Policy Pract. 133, 82–94 (2020)
Kumar Mitra, S.: Impact of carsharing on the mobility of lower-income populations in California. Travel
Behav. Soc. 24, 81–94 (2021). https://​doi.​org/​10.​1016/j.​tbs.​2021.​02.​005
Lane, C.: PhillyCarShare: first-year social and mobility impacts of carsharing in Philadelphia. Pennsylvania.
Transp. Res. Rec. (2005). https://​doi.​org/​10.​3141/​1927-​18
 Transportation

Le Vine, S., Polak, J.: The impact of free-floating carsharing on car ownership: early-stage findings from
London. Transp. Policy 75, 119–127 (2019). https://​doi.​org/​10.​1016/j.​tranp​ol.​2017.​02.​004
Li, L., Zhang, Y.: An extended theory of planned behavior to explain the intention to use carsharing: a
multi-group analysis of different sociodemographic characteristics. Transportation (2021). https://​doi.​
org/​10.​1007/​s11116-​021-​10240-1
Li, W., Kamargianni, M.: An integrated choice and latent variable model to explore the influence of attitudi-
nal and perceptual factors on shared mobility choices and their value of time estimation. Transp. Sci.
54, 62–83 (2020). https://​doi.​org/​10.​1287/​trsc.​2019.​0933
Liao, F., Correia, G.: Electric carsharing and micromobility: a literature review on their usage pattern,
demand, and potential impacts. Int. J. Sustain. Transp. 16, 269–286 (2022)
Liao, F., Molin, E., Timmermans, H., van Wee, B.: Carsharing: the impact of system characteristics on its
potential to replace private car trips and reduce car ownership. Transportation 47, 935–970 (2020)
Likert, R.: A technique for the measurement of attitudes. Arch. Psychol. (1932)
Luna, T., Uriona-Maldonado, M., Silva, M., Vaz, C.: The influence of e-carsharing schemes on electric
vehicle adoption and carbon emissions: an emerging economy study. Transp. Res. Part D: Transp.
Environ. (2020). https://​doi.​org/​10.​1016/j.​trd.​2020.​102226
Luo, W., Sun, L., Wang, S., Rong, J.: Travel choice of car-sharing based on Lewin metal of behavior. J. Bei-
jing Univ. Technol. 45, 476–484 (2019). https://​doi.​org/​10.​11936/​bjutx​b2018​020011
Martin, E., Shaheen, S.: Greenhouse gas emission impacts of carsharing in North America. IEEE Trans.
Intell. Transp. Syst. 12, 1074–1086 (2011a). https://​doi.​org/​10.​1109/​TITS.​2011.​21585​39
Martin, E., Shaheen, S.: The impact of carsharing on public transit and non-motorized travel: an exploration
of North American carsharing survey data. Energies 4, 2094–2114 (2011b). https://​doi.​org/​10.​3390/​
en411​2094
Martin, E., Shaheen, S.: Impacts of Car2Go on vehicle ownership, modal shift, vehicle miles traveled, and
greenhouse gas emissions: an analysis of five North American cities. Technical Report UC Berkeley
Transportation Sustainability Research Center (2016)
Migliore, M., D’Orso, G., Caminiti, D.: The environmental benefits of carsharing: the case study of
Palermo. Transp. Res. Procedia 48, 2127–2139 (2020). https://​doi.​org/​10.​1016/j.​trpro.​2020.​08.​271
Miljure, A.K., Ben: ShareNow, formerly Car2Go, to end service in North America (2019). https://​bc.​ctvne​
ws.​ca/​share-​now-​forme​rly-​car2go-​to-​end-​servi​ce-​in-​north-​ameri​ca-1.​47355​82?​cache=%​3Fcli​pId%​
3D685​97. Accessed 11 Apr 2024
Mokhtarian, P.L., Salomon, I., Redmond, L.S.: Understanding the demand for travel: it’s not purely
‘derived’. Innov. Eur. J. Soc. Sci. Res. 14, 355–380 (2001)
Møller, T., Simlett, J.: Micromobility: moving cities into a sustainable future. Technical Report EY (2020)
Monteiro, M.M., Azevedo, C.M.L., Kamargianni, M., Shiftan, Y., Gal-Tzur, A., Tavory, S.S., Antoniou, C.,
Cantelmo, G.: Car-sharing subscription preferences: the case of Copenhagen, Munich, and Tel Aviv-
Yafo (2022). https://​doi.​org/​10.​48550/​arXiv.​2206.​02448
Monteiro, M.M., Lima de Azevedo, C.M., Kamargianni, M., Cantelmo, G., Shoshany Tavory, S., Gal-Tzur,
A., Antoniou, C., Shiftan, Y.: Car-sharing subscription preferences and the role of incentives: the case
of Copenhagen, Munich, and Tel Aviv-Yafo. Case Stud. Transp. Policy 12, 101013 (2023). https://​doi.​
org/​10.​1016/j.​cstp.​2023.​101013
Moreau, H., de Jamblinne de Meux, L., Zeller, V., D’Ans, P., Ruwet, C., Achten, W.M.: Dockless e-scooter:
a green solution for mobility? Comparative case study between dockless e-scooters, displaced trans-
port, and personal e-scooters. Sustainability 12, 1803 (2020)
Morwinsky, S.: Car-sharing—North America. Technical Report Statista (2023). https://​www.​stati​sta.​com/​
outlo​ok/​mmo/​shared-​mobil​ity/​car-​shari​ng/​north-​ameri​ca. Accessed 11 Apr 2024
Mundaca, L., Román-Collado, R., Cansino, J.M.: Assessing the impacts of social norms on low-carbon
mobility options. Energy Policy 162, 112814 (2022a)
Mundaca, L., Zhu, X., Hackenfort, M.: Behavioural insights for sustainable energy use (2022b)
Münzel, K., Boon, W., Frenken, K., Blomme, J., van der Linden, D.: Explaining carsharing supply across
western European cities. Int. J. Sustain. Transp. 14, 243–254 (2020)
Namazu, M., Dowlatabadi, H.: Characterizing the GHG emission impacts of carsharing: a case of Vancou-
ver. Environ. Res. Lett. (2015). https://​doi.​org/​10.​1088/​1748-​9326/​10/​12/​124017
Namazu, M., MacKenzie, D., Zerriffi, H., Dowlatabadi, H.: Is carsharing for everyone? Understanding the
diffusion of carsharing services. Transp. Policy 63, 189–199 (2018)
Nansubuga, B., Kowalkowski, C.: Carsharing: a systematic literature review and research agenda. J. Serv.
Manag. 32, 55–91 (2021)
Narayanan, S., Antoniou, C.: Expansion of a small-scale car-sharing service: a multi-method framework
for demand characterization and derivation of policy insights. J. Transp. Geogr. 104, 103438 (2022).
https://​doi.​org/​10.​1016/j.​jtran​geo.​2022.​103438
Transportation

Nijland, H., van Meerkerk, J.: Mobility and environmental impacts of car sharing in the Netherlands. Envi-
ron. Innov. Soc. Trans. 23, 84–91 (2017). https://​doi.​org/​10.​1016/j.​eist.​2017.​02.​001
Olaru, D., Greaves, S., Leighton, C., Smith, B., Arnold, T.: Peer-to-Peer (P2P) carsharing and driverless
vehicles: attitudes and values of vehicle owners. Transp. Res. Part A: Policy Pract. 151, 180–194
(2021). https://​doi.​org/​10.​1016/j.​tra.​2021.​07.​008
Pearson, K.: X. On the criterion that a given system of deviations from the probable in the case of a correlated
system of variables is such that it can be reasonably supposed to have arisen from random sampling.
Lond. Edinb. Dublin Philos. Mag. J. Sci. 50, 157–175 (1900). https://​doi.​org/​10.​1080/​14786​44000​94638​
97
Poltimäe, H., Rehema, M., Raun, J., Poom, A.: In search of sustainable and inclusive mobility solutions for rural
areas. Eur. Transp. Res. Rev. (2022). https://​doi.​org/​10.​1186/​s12544-​022-​00536-3
Pourhashem, G., Malichová, E., Piscová, T., Kováčková, T.: Gender difference in perception of value of travel
time and travel mode choice behavior in eight European countries. Sustainability (2022). https://​doi.​org/​
10.​3390/​su141​610426
Pronello, C., Gaborieau, J.-B.: Engaging in pro-environment travel behaviour research from a psycho-social
perspective: a review of behavioural variables and theories. Sustainability 10, 2412 (2018)
Pyzyk, K.: formerly Car2Go, decides to fully exit North America (2019). https://​www.​smart​citie​sdive.​com/​
news/​share-​now-​forme​rly-​car2go-​decid​es-​to-​fully-​exit-​north-​ameri​ca/​564104/
Queiroz, M., Celeste, P., Moura, F.: School commuting: the influence of soft and hard factors to shift to public
transport. Transp. Res. Procedia 47, 625–632 (2020). https://​doi.​org/​10.​1016/j.​trpro.​2020.​03.​140
R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Comput-
ing Vienna, Austria (2023). https://​www.R-​proje​ct.​org/
Rabbitt, N., Ghosh, B.: A study of feasibility and potential benefits of organised car sharing in Ireland. Transp.
Res. Part D: Transp. Environ. 25, 49–58 (2013). https://​doi.​org/​10.​1016/j.​trd.​2013.​07.​004
Rahimi, A., Azimi, G., Jin, X.: Examining human attitudes toward shared mobility options and autonomous
vehicles. Transport. Res. F: Traffic Psychol. Behav. 72, 133–154 (2020a)
Rahimi, A., Azimi, G., Jin, X.: Examining human attitudes toward shared mobility options and autonomous
vehicles. Transport. Res. F: Traffic Psychol. Behav. 72, 133–154 (2020b). https://​doi.​org/​10.​1016/j.​trf.​
2020.​05.​001
Raux, C., Zoubir, A., Geyik, M.: Who are bike sharing schemes members and do they travel differently? The
case of Lyon’s “Velo’v’’ scheme. Transp. Res. Part A: Policy Pract. 106, 350–363 (2017)
Raveau, S., Álvarez-Daziano, R., Yáñez, M.F., Bolduc, D., de Dios Ortúzar, J.: Sequential and simultaneous
estimation of hybrid discrete choice models: some new findings. Transp. Res. Rec. 2156, 131–139 (2010)
Rayle, L., Dai, D., Chan, N., Cervero, R., Shaheen, S.: Just a better taxi? A survey-based comparison of taxis,
transit, and ridesourcing services in San Francisco. Transp. Policy 45, 168–178 (2016)
Reck, D.J., Axhausen, K.W., Hensher, D.A., Ho, C.Q.: Multimodal transportation plans: empirical evidence
on uptake, usage and behavioral implications from the Augsburg Maas trial. In: 100th Annual Meeting
of the Transportation Research Board (TRB 2021)(virtual) (pp. TRBAM–21). IVT, ETH Zurich (2021)
Redmond, L.: Identifying and analyzing travel-related attitudinal, personality, and lifestyle clusters in the San
Francisco Bay Area. Master’s Thesis UC Davis: Institute of Transportation Studies (2000). https://​escho​
larsh​ip.​org/​uc/​item/​0317h​7v4
Revelle, W.: Experimental approaches to the study of personality. In: Handbook of Research Methods in Per-
sonality Psychology, pp. 37–61. The Guilford Press (2007)
Rogers, E.M., Singhal, A., Quinlan, M.M.: Diffusion of innovations. In: An Integrated Approach to Communi-
cation Theory and Research, pp. 432–448. Routledge (2014)
Schlüter, J., Weyer, J.: Car sharing as a means to raise acceptance of electric vehicles: an empirical study on
regime change in automobility. Transp. Res. F: Traffic Psychol. Behav. 60, 185–201 (2019). https://​doi.​
org/​10.​1016/j.​trf.​2018.​09.​005
Schure, J., Napolitan, F., Hutchinson, R.: Cumulative impacts of carsharing and unbundled parking on vehicle
ownership and mode choice. Transp. Res. Rec. (2012). https://​doi.​org/​10.​3141/​2319-​11
Seo, J., Lee, S.: Who gives up a private car for a car-sharing service? An empirical case study of Incheon City.
South Korea. Int. J. Sustain. Transp. (2021). https://​doi.​org/​10.​1080/​15568​318.​2021.​19490​77
Shaheen, S., Cohen, A.: Innovative mobility: carsharing outlook carsharing market overview, analysis, and
trends. Technical Report Institute of Transportation Studies at UC Berkeley (2020). https://​escho​larsh​ip.​
org/​uc/​item/​9jh43​2pm
Shaheen, S., Cohen, A., Farrar, E.: Carsharing’s impact and future. Adv. Transp. Policy Plan. 4, 87–120 (2019).
https://​doi.​org/​10.​1016/​bs.​atpp.​2019.​09.​002
Shaheen, S., Martin, E.: Unraveling the modal impacts of Bikesharing. ACCESS Magazine, p. 9 (2015)
 Transportation

Shaheen, S., Martin, E., Bansal, A.: One-way electric vehicle carsharing in San Diego: an exploration of the
behavioral impacts of pricing incentives on operational efficiency. Technical Report UC Berkeley Trans-
poratation Sustainability Research Center (2018). https://​doi.​org/​10.​7922/​G22Z1​3P5
Stasko, T., Buck, A., Oliver Gao, H.: Carsharing in a University setting: impacts on vehicle ownership, parking
demand, and mobility in Ithaca. NY. Transp. Policy 30, 262–268 (2013). https://​doi.​org/​10.​1016/j.​tranp​
ol.​2013.​09.​018
Steg, L.: Sustainable transportation: a psychological perspective. IATSS Res. 31, 58–66 (2007)
Stillwater, T., Mokhtarian, P.L., Shaheen, S.A.: Carsharing and the built environment: geographic information
system-based study of one US operator. Transp. Res. Rec. 2110, 27–34 (2009)
Sun, S., Liu, Y., Yao, Y., Duan, Z., Wang, X.: The determinants to promote college students’ use of car-sharing:
an empirical study at Dalian Maritime University. China. Sustainability (2021). https://​doi.​org/​10.​3390/​
su131​26627
Te, Q., Lianghua, C.: Carsharing: mitigation strategy for transport-related carbon footprint. Mitig. Adapt. Strat.
Glob. Change 25, 791–818 (2020)
Temme, D., Paulssen, M., Dannewald, T.: Integrating latent variables in discrete choice models: how higher-
order values and attitudes determine consumer choice. Technical Report SFB 649 Discussion Paper
(2007)
Thurner, T., Fursov, K., Nefedova, A.: Early adopters of new transportation technologies: attitudes of Russia’s
population towards car sharing, the electric car and autonomous driving. Transp. Res. Part A: Policy
Pract. 155, 403–417 (2022). https://​doi.​org/​10.​1016/j.​tra.​2021.​11.​006
Tilley, S., Houston, D.: The gender turnaround: young women now travelling more than young men. J. Transp.
Geogr. 54, 349–358 (2016). https://​doi.​org/​10.​1016/j.​jtran​geo.​2016.​06.​022
Tirachini, A., Chaniotakis, E., Abouelela, M., Antoniou, C.: The sustainability of shared mobility: Can a
platform for shared rides reduce motorized traffic in cities? Transp. Res. Part C: Emerg. Technol. 117,
102707 (2020). https://​doi.​org/​10.​1016/j.​trc.​2020.​102707
Tirachini, A., del Río, M.: Ride-hailing in Santiago de Chile: users’ characterisation and effects on travel behav-
iour. Transp. Policy 82, 46–57 (2019)
Vermeulen, B., Goos, P., Vandebroek, M.: Models and optimal designs for conjoint choice experiments includ-
ing a no-choice option. Int. J. Res. Mark. 25, 94–103 (2008)
Vij, A., Walker, J.L.: How, when and why integrated choice and latent variable models are latently useful.
Transp. Res. Part B: Methodol. 90, 192–217 (2016)
von Wieding, S., Sprei, F., Hult, C., Hult, Å., Roth, A., Persson, M.: Drivers and barriers to business-to-business
carsharing for work trips–a case study of Gothenburg. Sweden. Case Stud. Transp. Policy 10, 2330–2336
(2022)
Watanabe, C., Naveed, K., Neittaanmäki, P., Fox, B.: Consolidated challenge to social demand for resilient plat-
forms-lessons from Uber’s global expansion. Technol. Soc. 48, 33–53 (2017)
Wu, C., Le Vine, S., Clark, M., Gifford, K., Polak, J.: Factors associated with round-trip carsharing frequency
and driving-mileage impacts in London. Int. J. Sustain. Transp. 14, 177–186 (2020). https://​doi.​org/​10.​
1080/​15568​318.​2018.​15384​01
Xu, J.: Generational trends of gendered mobility: How do they interact with geographical contexts? J. Transp.
Geogr. 82, 102623 (2020)
Yang, H., Ma, Q., Wang, Z., Cai, Q., Xie, K., Yang, D.: Safety of micro-mobility: analysis of e-scooter crashes
by mining news reports. Accid. Anal. Prev. 143, 105608 (2020)
Yazdanpanah, M., Hosseinlou, M.: The influence of personality traits on airport public transport access mode
choice: a hybrid latent class choice modeling approach. J. Air Transp. Manag. 55, 147–163 (2016).
https://​doi.​org/​10.​1016/j.​jairt​raman.​2016.​04.​010
Ye, J., Wang, D., Li, X., Axhausen, K., Jin, Y.: Assessing one-way carsharing’s impacts on vehicle ownership:
evidence from Shanghai with an international comparison. Transp. Res. Part A: Policy Pract. 150, 16–32
(2021). https://​doi.​org/​10.​1016/j.​tra.​2021.​05.​012
Zhang, Y., Li, L.: Intention of Chinese college students to use carsharing: an application of the theory of
planned behavior. Transp. Res. F: Traffic Psychol. Behav. 75, 106–119 (2020). https://​doi.​org/​10.​1016/j.​
trf.​2020.​09.​021
Yoon, T., Cherry, C.R., Jones, L.R.: One-way and round-trip carsharing: a stated preference experiment in Bei-
jing. Transp. Res. Part D: Transp. Environ. 53, 102–114 (2017)
Zhang, R., Spieser, K., Frazzoli, E., Pavone, M.: Models, algorithms, and evaluation for autonomous mobility-
on-demand systems. In: 2015 American Control Conference (ACC), pp. 2573–2587. IEEE (2015)

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.