Mobility-On-Demand: An Empirical Study of Internet-Based Ride-Hailing Adoption Factors, Travel Characteristics and Mode Substitution Effects

Authors accepted version of original article to appear in Transportation Research Part C: Emerging Technologies
Mobility-on-demand: An empirical study of internet-based ride-hailing

adoption factors, travel characteristics and mode substitution effects
To cite: Acheampong, R. A., Siiba, A., Okyere, D. K., & Prosper Tuffour, J. (Accepted/In press).
Mobility-on-demand: An empirical study of internet-based ride-hailing adoption factors, travel
characteristics and mode substitution effects. Transportation Research Part C: Emerging
Technologies.
Corresponding Author:
Ransford A. Acheampong [email protected]
Department of Planning and Environmental Management, University of Manchester
Abstract
Ride-hailing services are shaping travel behaviours and emergent urban mobility patterns.
From their initial diffusion centres in North America and Europe, these on-demand mobility
services are increasingly becoming available in developing countries. Yet, empirical research
from these contexts on the impact of ride-hailing services is lacking. To address this gap, this
paper examines the factors driving the adoption of ride-hailing and the associated travel
characteristics and mode substitution effects in Ghana, Sub-Saharan Africa. Using data from
a large sample survey (N = 1,188) of commuters in a multi-variable structural equation model,
the paper shows that socio-demographic factors, perceived benefits and ease of use of ride-
hailing, perceived safety risks and car-dependent lifestyles influence adoption and use of ride-
hailing services. Similar to other contexts, individuals’ reference ride-hailing trips were mainly
for ‘special occasion’ purposes (51%), but work and school travels were also high (41%).
Shorter travel times (≤30 minutes) and single passenger journeys within inner-suburban and
outer-suburban localities typify ride-hailing trips. This contrasts with other contexts where
ride-hailing is used frequently by urban dwellers and less so by those in the suburbs. Ride-
hailing use replaced conventional taxis (51%), public transport (36%), private car (10%) and
walking (1%), suggesting mode substitution effects for individuals’ reference trips. Further
exploration of a full day’s travel mode choices also revealed that individuals use other
available modes of transport in addition to ride-hailing services. However, multi-modal
integration is weak, suggesting that ride-hailing tends to be used alone for full door-to-door
journeys, instead of complementing other existing modes in serving first/last mile access for
example. The implications of the findings for sustainable mobility are discussed.
Key words: Mobility-on-demand; platform mobility; ride-hailing; travel behaviour; mode-

substitution; Ghana
1
1 Introduction
On-demand platform mobility services are increasingly becoming common in urban areas.
The growth of these services in recent years has in part been enabled by innovation in
Information and Communication Technology (ICT) that is being leveraged by Transportation
Network Companies (TNCs) to offer flexible and convenient mobility solutions. For example,
ICT has enabled the implementation of station-based and free-floating car-sharing, bike-
sharing and (e)scooter-sharing (see e.g. Degele et al., 2018; Shaheen and Cohen, 2013; Park
et al., 2014). These existing on-demand mobility services are also evolving into new ones, such
as Mobility-as-a-Service (Maas) and connected autonomous transport.
Today, internet-based on-demand mobility services commonly known as ride-hailing are

spreading globally from their initial diffusion centres in North America and Europe. TNCs, such
as Uber and Lyft, are penetrating new markets including those in Latin America and Sub-
Saharan Africa. The presence of ride-hailing services is already shaping travel behaviours and
emergent urban mobility patterns. Initial exploratory research has found that, the availability
of ride-hailing services on the one hand, can induce new trips (Rayle et al., 2016) and take
away travel demand from existing public transit service overtime (Doppelt, 2018; Lindsay,
2017). Some have concluded to the contrary that ride-hailing appears to complement and
increase public transport ridership by addressing the first and the last mile problem in
passenger transport (Circella and Alemi, 2018; Hall et al., 2018; Sadowsky and Nelson, 2017).
The transformational nature of ride-hailing services has been widely acknowledged. However,
the data required to assess their impacts and to aid travel demand management strategies
are not being made readily available by TNCs. Consequently, to-date, a number of empirical
research, relying on census data or large sample surveys have sought to explore the impacts
of ride-hailing services. Specifically, such research largely comes from North America (see e.g.
Ward et al., 2019; Grahn et al., 2019; Graehler et al., 2019; Dias et al., 2019; Lavieri and Bhat,
2019; Yan et al., 2019; Brown, 2018; Alemi et al., 2018b; Conway et al., 2018; Clewlow and
Mishra, 2017; Dias et al., 2017; Hoffmann et al., 2016; American Public Transportation
Association, 2016; Anderson, 2014). A limited number of studies have also explored
governance, policy and mobility issues around emerging ride-haling services in other regions,
including Asia (Dong et al., 2018; Nie, 2017); South America (Haddad, et al., 2019; Gomez-
Morantes et al., 2019) and Sub-Saharan Africa (Agyemang, 2020; Giddy, 2019).
Given that ride-hailing services are quite new and evolving, evidence regarding their wider
impacts is still emerging and therefore inconclusive. In developing countries, there is a dearth
of empirical evidence in the literature that would aid our understanding of critical issues
around ride-hailing, including the factors driving adoption, the associated travel behaviours
and mode complementarity and substitution effects. This study contributes to filling these
gaps by examining ride-hailing adoption factors, travel characteristics and mode substitution
effect in Ghana, Sub-Saharan Africa, where TNCs including Uber and Bolt/Taxify have been
operating rid-hailing services in the last few years.
2
The main contributions of the study are as follows. Firstly, the paper draws on insights from
initial exploratory studies to assemble and deploy a theory-grounded conceptual framework
to model the determinants of ride-hailing adoption. The framework presented here would
therefore be useful for further empirical studies in other contexts. Moreover, like many
developing countries, the Ghana context is particularly unique in terms of the existing
transport regimes prior to ride-hailing, as well as the quality of public transport and
conventional taxi services that precede and co-exist with ride-hailing. Thus, the current paper
provides detailed empirical insights that do not only corroborate findings from previous
research, but they also offer nuanced understanding of the ways in which ride-hailing is
shaping travel behaviour in previously unexplored contexts. Lastly, we outline and reflect on
the implications of the findings regarding the factors driving adoption of ride-hailing services
and actual use patterns of this new form of mobility.
2 The rise of platform on-demand mobility services

Ride-hailing services leverage ICT and the widespread adoption of smartphones. These
mobility services also form part of wider changes in everyday consumption practices where
consumers, including commuters, replace personal ownership of cars with access-based
alternatives such as shared-mobility services (Barnes and Mattson, 2016; Bardhi and
Eckhardt, 2012; Burkhardt and Millard-Ball, 2006). Using online mobile applications, service
providers match passengers who have requested a ride with drivers near the intended trip’s
origin (Conway et al., 2018). Typically, an individual could travel alone or where ride-sharing
options are available, couple with others with whom they share the same destination (Saadi
et al., 2017).
People prefer ride-hailing services because they are convenient to use and can be affordable
compared with other options (Pham et al., 2017). Moreover, with ride-hailing, people can
avoid driving while intoxicated, travel time could be shorter and individuals could avoid
having to use personal cars for trips to busy urban centres where parking is limited (Lavieri
and Bhat, 2019). The emerging evidence show that ride-hailing is commonly used for social
and recreational trips (Hampshire et al., 2018; Rayle et al., 2016), for errands and work travel
Lavieri et al.,(2018), and on special occasions (Grahn et al., 2019). Similar to other ICT-enabled
mobility services (see e.g. Acheampong and Siiba, 2019; Clewlow and Mishra, 2017; Prieto et
al., 2017; Juschten et al., 2019; Efthymiou et al., 2013), users of ride-hailing services tend to
be younger, highly educated with relatively higher incomes (Grahn et al., 2019; Dias et al.,
2019; Lavieri and Bhat, 2019; Schaller, 2018).
Studies on the impacts of on-demand ride-hailing services are fraught with particular
challenges including the relative newness of the phenomenon itself and the fact that data on
usage is not readily available from service providers. Notwithstanding these challenges, a
growing body of research has attempted to examine the impacts of ride-hailing mobility
services. The emerging evidence points to both positive and negative impacts. On the one
hand, ride-hailing has been found to: replace private car use, thereby reducing parking
3
demand (Henao and Marshall, 2019), reduce the overall amount of driving per person (Rayle
et al., 2016) and increase public transport ridership by addressing the first and the last mile
problem in passenger transport (Circella and Alemi, 2018; Sadowsky and Nelson, 2017). On
the other hand, the evidence also suggest that ride-hailing services influence travel behaviour
by inducing new trips, thereby changing mobility patterns in urban areas (Dias et al., 2019;
Alemi et al., 2018a; Rayle et al., 2016). For example, Alemi et al., (2018a) and Rayle et al.,
(2016) found that overall, about eight percent of the users in their respective studies in San
Francisco, USA, would not have travelled, at all, if ride-hailing services were not available. It
is worth mentioning that these studies have, so far, been largely undertaken in the context of
a few cities in North America. Thus, it would be useful to explore how this new form of
mobility is shaping emergent mobility patterns and the underlying choice behaviours in
different contexts.
3 Understanding ride-hailing adoption—A conceptual model and underlying

theories
Comprehensive theoretical models have hardly underpinned studies carried out so far on
ride-hailing adoption factors, probably because of the exploratory nature of these initial
efforts. Notwithstanding, they provide rich empirical insights that could form the basis for
assembling a conceptual model in subsequent research. At the same time, research on
technology and innovation acceptance and diffusion, and choice behaviour under volitional
control do provide various theoretical models that could be adapted to study the adoption of
ride-hailing services. This study advances such a conceptual model by integrating relevant
concepts from the Technology Acceptance Model (TAM) Davis et al., 1989), Theory of Planned
Behaviour (TPB) (Ajzen, 1991),Technology Diffusion Theory (TDT) (Rogers, 1962, 2000) as well
as the emerging evidence on how the internet, mobile devices and applications are shaping
travel behaviour (see e.g. Ettema, 2018; Schwanen, 2015).
Fig 1: Conceptual model of on-demand ride-hailing mobility service adoption

NB: Double-head arrows show hypothesized association effects while single-headed arrows show direct effects
4
As illustrated in Fig 1, the proposed conceptual model integrates relevant concepts from the
original Technology Acceptance Model (TAM) (see Davis et al., 1989; Venkatesh and Davis,
2000). TAM posits that acceptance of new technology is partly influenced by individuals’
perception of its benefits and ease of use. Perceived benefits reflect an individual’s belief of
the extent to which adopting a new technology will enhance the performance of specific tasks
or activities, while perceived ease of use captures an individual’s belief about the extent to
which interacting with the new technology will be free of effort. Across a wide range of
applications, including the adoption of new mobility services, the evidence suggest that
perceived usefulness and ease of use tend to be positively related. Each of these factors, in
turn, has direct influence on adoption intentions or actual adoption and use of new
technologies (see e.g. Acheampong and Cugurullo, 2019; Wang et al., 2018; Park et al., 2014;
Hubert et al., 2017; Tan et al. 2014 ; Cheung and Vogel, 2013; Venkatesh and Davis, 2000). In
the conceptual model, two types of benefits of ride-hailing are identified: The first is
perceived instrumental/hedonic benefits, which reflect the expected utilitarian advantages
of ride-hailing as an alternative travel mode, such as convenience, reliability and travel time
savings. The second category of benefits identified, relates to individuals’ perception of the
advantages or otherwise that using ride-hailing services could have for the environment.
Perceived benefits and ease of use of ride-hailing are hypothesized to covary, with each of
them also having direct influence on ride-hailing adoption and use.
In addition, two concepts are drawn from the Theory of Planned Behaviour (TPB) (Ajzen,
1991) into our conceptual model. The first is attitude, which refers to an individual’s
expectation of the outcomes of an action or behaviour and the personal values that are
attached to them (Ajzen, 1991; Sutton et al., 2003). Indeed, ride-hailing services provide car-
based transport solutions. This implies that individuals’ attitude towards car ownership and
use would likely influence adoption (Circella and Alemi, 2018). From this theoretical and
empirical understanding, we capture a range of attitudinal variables to explain why people
use ride-hailing. Firstly, both the instrumental/hedonic and environmental-related attitudes
regarding car-ownership and use are identified. These attitudinal variables are hypothesized
to covary with each other. Car ownership attitudes are also posited to correlate with
perceived benefits of ride-hailing. The framework also captures car-dependent lifestyles,
which is expected to correlate with car ownership attitudes and perceived benefits of ride-
hailing. Each of these attitudinal factors are hypothesized to directly influence ride-hailing
adoption.
Furthermore, travel choices have environmental consequences. Indeed, environmental

awareness has grown and people are increasingly becoming concerned about the
environmental impacts of daily mobility choices. Environmental concerns have been found to
be associated with mobility choices (Acheampong and Siiba, 2019; Acheampong and
Cugurullo, 2019; Clark et al., 2016; Hess et al., 2012; Flamm, 2009). For example, Flamm
(2009) found that households with pro-environmental attitudes own fewer and more fuel-
efficient vehicles, while Clarke et al., (2016) showed that pro-environmental attitude
5
precedes switch from car-based transport. In the light of the aforementioned evidence, we
hypothesize that environmental attitudes and values will correlate with car-ownership
attitude, car-dependency and perceived benefits of ride-hailing services, as well as influence
directly the use of this new form of mobility.
New on-demand mobility services, such as ride-hailing depend on advances in technology and
ICT and the evidence suggest that the internet, mobile devices and applications are having
considerable impacts on travel behaviour (Okyere et al., 2018; Ettema, 2018; Schwanen,
2015; Mokhtarian, 2002; Wang and Law, 2007). In view of this, the proposed conceptual
model includes factors that allow eliciting individuals’ attitude towards technology in general,
their openness to technological innovation and use of ICT in their daily mobility decisions. In
addition, Roger’s Technology Diffusion Theory (TDT) (Rogers, 1962, 2000) has shown that the
innovation characteristics of a new product or technology as perceived by would-be users, is
crucial to its adoption. Thus, engaging the theoretical insight from TDT, our conceptual model
also integrates perceived innovativeness of ride-hailing as a distinct explanatory variable. In
the conceptual model, each of these technology and innovation-related factors is
hypothesized to covary with the other, and to directly influence ride-hailing use.
Another related concept that is considered relevant is perceived behavioural control.

Originally included in Ajzen’s Theory of Planned Behaviour (Ajzen, 1991), this concept is used
in the context of the conceptual model for this study to capture individuals’ perception of
their confidence in being able to use ride-hailing mobile applications—an important
determinant of actually being able to use this form of on-demand mobility service. We assume
that perceived confidence and perceived ease of use will covary and that each of these factors
will have direct explanatory effect on ride-hailing use. Moreover, using digital technologies
and services come with security and privacy risks (Wang et al., 2019; Hong, 2017; Lee et al.,
2017; Edelman and Geradin, 2015; Rogers, 2015; Featherman and Pavlou, 2003). In view of
this, the conceptual model also captures individuals’ perception of the safety risks associated
with using ride-hailing services as another determining factor of adoption. Perceived safety
risk of ride-hailing use is assumed to correlate with perceived confidence in being able to use
these services. These factors, in turn, are expected to directly influence whether or not
individuals use this new form of mobility.
Finally, socio-demographic factors, including age, gender, income and education are captured
in the conceptual model. We expect socio-demographic factors to correlate with the latent
variables identified in the conceptual model and to directly explain ride-hailing use. Given the
multiplicity of possible relationships between socio-demographic factors and other variables,
we do not map them out in the conceptual model. Instead, we demonstrate these effects
later in section 5.2.3 where the results of the statistical modelling of the factors underpinning
ride-hailing adoption and use are presented.
In the next section, the methodology for utilising the proposed conceptual model to
empirically examine ride-hailing adoption and use patterns is outlined.
6
4 Methodology
4.1 Study context
Ride-hailing services have become available in Ghana since Uber first operated services in
Accra, the country’s capital and biggest city around 2016. Uber’s main local competitor is Bolt
formerly known as Taxify, an Estonian Transportation Network Company (TNC) operating in
other Sub-Saharan African countries, including Tanzania, Nigeria and South Africa. Aside from
Uber and Bolt, a number of relatively smaller TNCs have emerged in the last couple of years
in Ghana’s major cities, including Yenko Ghana, Urughana, Accra Cab, and Poki Cab. Alongside
the services provided by the TNCs, conventional taxis licenced by local governments operate
‘station-based’1 and ‘floating’2 services in towns and cities (see Kufuor, 2019; Simmons, 2018;
Hart, 2016).
Using traditional taxi services in Ghana costs more than using public transport services known
locally as Trotro3. Taxis therefore tend to attract more people with relatively higher incomes
who do not own cars or are used less frequently for incidental trips by people in different
socio-economic groups (Acheampong, 2020; Agyemang, 2017). Most of the urban population
use Trotro/Minibus, which is known for its overall inferior service quality and lax safety
standards (Birago et al., 2017; Esson et al., 2016; Amoh-Gyimah and Aidoo, 2013).
Thus, the entry of TNCs providing ride-hailing services does not only constitute an
unprecedented innovation in the urban passenger transport sector, but also a major
disruption to the existing taxi and public transport regimes in Ghana. While TNCs and their
ride-hailing services are now part of the urban transportation system, the existing transport
governance frameworks in Ghana are currently lagging behind. Originally enacted to govern
conventional taxi operations, the existing regulatory regime seem inadequate to regulate
TNCs and their ride-hailing services (Agyemang, 2020). Ride-hailing vehicles are not licenced
to provide taxi services and they do not follow basic regulatory requirements including being
distinctly marked and belonging to a recognized unionized group as required of conventional
taxis (Agyemang, 2020). It is within this context that the adoption factors, travel behaviours
around new ride-hailing services and their implications for transport planning and policy are
examined, using survey data collected from commuters in Accra and Kumasi. The next
sections describe the survey questionnaire design and the data collection methods.
1
Taxis operate from dedicated taxi ranks in the urban area, where taxi drivers wait their turn for passengers to
board.
2
Unlike station-based taxi service, floating taxis roam the urban area pulling over by the roadside to pick any
passengers that hail them to stop. It is worth noting that drivers registered to stations do sometimes ‘float’,
prospecting for passengers, especially during off-peak periods where taxi ranks have fewer customers.
3
In Ghana, Trotro are mainly 10-19 seater privately operated minibuses used as the main form of public
transport. While Trotro services are operated by Transport Unions comprising vehicle owners and/or drivers,
they are regulated by local governments. Buses, which are another form of public transport are different from
Trotro in that they are large capacity vehicles. Bus services such as the BRT system in Accra are operated by the
state but some bus services are privately operated.
7
4.2 Survey questionnaire design

The survey questionnaire translated the factors presented in the conceptual model in Fig 1
into questions for the study participants. The questionnaire had five main parts outlined as
follows: Part I elicited the background socio-demographic characteristics of the survey
respondents as well as their locality of residence within their respective cities. It was expected
that not all the respondents would be familiar with on-demand platform mobility services in
general and ride-hailing services in particular and how they work. This would be the case,
especially among individuals who had never heard about and/or used any of the available
ride-hailing services. For this reason, in Part II of the questionnaire, the concept of ride-hailing
was introduced to the respondents in a preamble which included the following:
“..Ride-hailing has recently become available in major cities in Ghana. There are currently two
major services in Accra and Kumasi, provided by Uber and Bolt (formerly Taxify). Using ride-
hailing is similar to using a taxi. The only difference is that you need an App on your phone or
computer with internet access. You can book a ride-hailing service using the App when you
need to travel. A driver receives your booking or call and drives the car to pick you up and take
you where you are going. When you get to your destination, you pay a fare that is charged
based on the time you have been in the car…”
Following the above information, indicator items measuring perceived benefits, ease of use,
behavioural control, perceived innovation characteristics of ride-hailing and perceived safety
risks were presented to the respondents to evaluate on a five-point Likert Scale, ranging from
1(strongly disagree) to 5(strongly agree). All the survey respondents responded to these
items, irrespective of having used ride-hailing services or not. Subsequently, the survey
respondents also indicated whether they have used any of the ride-hailing services available
in their respective cities.
In Part III of the survey, the focus shifted to ride-hailing usage and the associated travel
characteristics and mobility patterns. Thus, only the respondents who had previously
indicated using a ride-hailing service in Part II of the survey responded to questions in Part III.
Here, the survey respondents indicated how long it had been, prior to taking the survey, since
they used a ride-hailing service. They also indicated the number of times they had done so in
the last seven days preceding the date of filling in the survey. Moreover, the respondents
stated, for their most recent journey using ride-hailing services, the trip purpose, time of day
of the trip and the specific reasons for ride-hailing. Regarding the reasons for ride-hailing, the
respondents could choose more than one reason that applied from an initial list presented to
them in the questionnaire. They could also include other reasons that were not provided in
the original list. Where the questionnaire statements implicitly involved comparison being
drawn with other modes, such as the statement ‘ride-hailing is a faster option’, individuals
were asked to indicate their responses relative to the other available public transport
alternatives in their city. For instance, in this particular example, individuals would indicate
the extent to which they considered ride-hailing as a faster travel option compared to the
conventional taxi and/or public transport (Trotro).
8
Further questions included in Part III of the survey allowed the respondents to provide more
information about their ride-hailing booking and the reference trip, including the waiting
time after booking; whether they travelled alone in the vehicle for the entire journey or they
shared the ride with others, and the estimated journey time and costs. The survey
respondents were also asked to indicate the locality of their reference ride-hailing trip origin
and destination to help better understand travel patterns.
Moreover, to understand mode substitution and complementarity effects, the survey

respondents were first asked to indicate what mode they typically used for the trips similar
to the one reported as their reference trip before ride-hailing was available. They also
subsequently indicated any other modes they used on the day of their referenced trip. The
rationale here was to examine mode substitution within the context of a full day’s travel
choices that could also involve using existing conventional modes.
In the penultimate part of the questionnaire, the survey explored whether ride-hailing was
used for school trips by respondents who have children or dependants of school going age
(i.e. Basic level to Senior High School). Thus, only adults with children, whether they have
themselves used ride-hailing or not, completed this part of the survey questionnaire. Those
who had booked ride-hailing services for their children’s most recent school journeys, were
asked to indicate the reasons for doing so, the estimated travel time and costs and origin and
destinations.
The final part of the questionnaire elicited further information from all the respondents
namely, their attitudes to the environment, car ownership and car use, and technology and
innovation. They also responded to questions formulated to elicit their openness to
innovation and use of ICT in travel, as captured in the conceptual model presented in Fig 1.
Statements measuring these concepts were presented to the respondents to evaluate on a
five-point Likert Scale ranging from 1(strongly disagree) to 5 (strongly agree). A grouping of
all the survey items are presented later under section five in Table 1 for socio-demographic
characteristics and Table 2 for the latent concepts captured in the conceptual model.
4.3 Data collection

The survey questionnaire was hosted online and could be self-completed by individuals who
could read and understand English and had access to the internet. For this group of
respondents, the research team reached them via various online and social media platforms,
including direct emailing, Facebook, WhatsApp and LinkedIn. To reach a wider target
audience, Individuals who received the questionnaire were encouraged to distribute it to
other individuals in their network. Mindful of the fact that only a segment of the population
could be reached online and would be able to self-complete the questionnaire, Field
Assistants also conducted face-to-face interviews, using versions of the questionnaire hosted
9
on Mobile Phones and Tablets. The personal interviews targeted relatively older populations
in the case study cities.
The survey was administered between May and August 2019. A total of 1,188 valid responses
from adults aged over 18 years was obtained. An estimated 65% of the respondents were
attracted via the online platforms while the remaining 35% of responses were obtained
through the personal interviews.
4.4 Structural equation modelling

As previously depicted in Fig 1, the conceptual model of ride-hailing adoption/use integrates
a number of variables that are considered latent— variables that are not directly observable,
but instead require the use of a number of indicator measures as presented later in Table 2
in the results section. Standard regression methods, such as linear regression models are not
suitable for the analysis of latent variables (Kline, 2016; Byrne, 2013). We therefore
considered Structural Equation Modelling (SEM) a suitable approach for our analysis.
SEM combines regression and factor analysis of latent variables to provide a robust statistical
approach to examine a set of qualitative causal hypotheses that are based on theory (see
Kline, 2016). The methodology involves specification of a Confirmatory Factor Analysis (CFA)
to test causal relationships between latent variables and regressing the latent variables on an
outcome variable. In our analysis, the outcome variable of interest is adoption of ride-hailing
services, while the latent explanatory variables are derived from the CFA of the relevant
Likert-scale based items that reflect the theoretical concepts/constructs in the conceptual
model, such as attitude and perceived benefits.
The SEM’s fit to ascertain the extent to which each of the conceptual model’s relationships
are supported by the empirical data was tested, using model identification indices including:
the Chi-square statistic (χ2) and Normed Chi-square (χ2/df); Normed Fit Index (NFI);
Comparative Fit Index (CFI); Tucker–Lewis Index (TLI); Incremental fit index (IFI); and Root
Mean Square Error of Approximation (RMSEA). NFI, CFI, TLI and IFI values ≥0.90 indicate
acceptable model fit, while RMSEA value of 0.01, 0.05, and 0.08 indicate excellent, good, and
mediocre fit, respectively (see Kline, 2016). In this paper, the SEM is estimated using AMOS
software.
5 Results
5.1 Background characteristics of respondents
Summary statistics on the socio-demographic characteristics of the survey respondents are
presented in Table 1. Out of the 1,188 valid responses, 51% and 49% resided in Accra and
Kumasi, respectively. Males constituted 52% of the total sample. The sample typifies the
population structure of Ghana and other Sub-Saharan African countries, which have a
10
relatively larger proportion of youthful population. The age-groupings of the survey

respondents have been demarcated to roughly correspond to the commonly used
generational cohorts as follows; 18-24 (Post-millennials), 25-39 (Millennials), 40-54
(Generation X) and 55+ (Baby-Boomers). Moreover, most of the respondents (63%) do not
own a car or have access to a private car. About 15% of the respondents own a car while the
remaining 22% do not own a car, but have access to at least one car belonging to their
household that they could occasionally use subject to availability. Additional descriptive
statistics on socio-demographic factors, including coupling, education, employment status
and income are presented in Table 1.
Table 1: Summary of characteristics of the survey respondents.

City of residence Accra: Kumasi 51% : 49%
Gender Female: Male 48%:52%
Age-groups 18-24 18%

25-39 53%
40-54 26%
55+ 3%
Coupling Couple: Single: Divorced/Widowed 45%:53%:2%
Have children? Yes: No 42%:58%
Children in school (KG- Yes: No 84%:16%

SHS)? (n =497)
Education Basic school (Primary + Junior High School) 16%

Senior High School (SH) 12%
Tertiary (undergraduate/Diploma)- currently enrolled 13%
Tertiary (undergraduate/Diploma)- completed 34%
Postgraduate (Master's degree or higher) 25%
Employment Employed (Full-time) 50%

Employed (Part-time) 18%
Retired 1%
Student (Full-time) 17%
Unemployed 14%
Income (n =812) ≤GH¢1,050 55%

GH¢1,051-2,000 19%
GH¢2,001-3,000 10%
GH¢3,001-4,000 5%
GH¢4,001-5,000 4%
≥GH¢5,001 7%
Car Ownership/Use I have a car 15%

I DON'T have my own car, but there is at least one car in my household that I 22%
can use if it is available
None of the above applies to me (don’t have a car) 63%
Driver’s Licensure Valid License: Learner, provisional: No 32%:5%:63%
11
5.2 Modelling ride-hailing adoption factors

In this section, a Structural Equation Model (SEM) is specified to examine the determinants
of ride-hailing adoption and use. Before specifying the model, summaries of the survey
responses regarding the latent explanatory variables and ride-hailing usage in the population
are presented in the sections that follow.
5.2.1 Descriptive summary of responses to latent concepts

Perceived benefits, ease of use, confidence and safety in using ride-hailing
Overall, a significant proportion of the survey respondents perceived ride-hailing to be
beneficial new mobility service (Table 2). Regarding perceived instrumental and hedonic
benefits, a minimum of 57% and a maximum of 74% of the survey respondents agreed that
ride-hailing would provide such benefits. This means majority of the respondents perceived
that ride-hailing would be flexible (66%), reliable (62%), relatively faster (57%) and convenient
(70%) travel mode option, compared to other available modes of transport. Moreover, 64%
and 74% of the respondents agreed that ride-hailing would be suitable for work and non-work
trip purposes, respectively. On the expected benefits related to the environment, more than
half of the survey respondents agreed that ride-hailing services would reduce the need for a
personal vehicle (52%), reduce congestion (61%) and potentially reduce travel-related
environmental pollution (52%). Following the description of ride-hailing services to the
respondents, most of them indicated that using the service would be easy for them (69%) and
that they felt very confident in being able to do so (74%). Regarding safety, more than half of
the respondents (56%) indicated that they would not feel safe travelling with strangers in a
ride-hailing vehicle, suggesting a lack of trust in internet-based ride-sharing mobility services
among most of the study participants.
Openness to innovation, technology attitude and ICT use in travel

Overall, most of the respondents indicated their readiness to use new technologies. About
73% of them agreed that they like to try things that are new and different while a little over
half of them indicated that they think they usually try new things before others (i.e. family
members and colleagues) do it. The role of external social environment was also
acknowledged, as half of the respondents agreed that they try new things (products and
services) when they find that other people are also using them. With specific reference to
ride-hailing, about 83% of the respondents agreed that this is an innovative mobility service
compared to the conventional taxi and public transport in their respective cities. Moreover,
regarding overall attitude towards technology, the survey results show that relatively fewer
respondents (i.e. 35%) disagreed that learning to use new technologies is often frustrating
compared to the 43% and 22% who agreed and were indifferent, respectively. The
importance to individuals of having Wi-Fi-internet access was indicated by 71% of the
respondents. About 73% also agreed that with their smartphones, moving around their city is
now easier than before, reflecting the increasing use of ICT in their daily travels.
12
Car ownership/use attitudes, car-dependency and environmental values

From the survey results, it becomes evident that the vast majority of the respondents favour
the expected instrumental and hedonic benefits associated with car ownership and use, such
as privacy (89%), comfort (86%), travel time (82%) and safety (83%)—Table 2. This attitude
towards the private car could be explained by the fact that public transport in urban Ghana is
generally of inferior quality and characterised by lax safety standards. Moreover, 66% of the
respondents agreed that car ownership confers social status. It is therefore not surprising that
about 82% indicated that they would definitely want to own a car. While the majority
favoured private car ownership, they also acknowledged that car-based transport pollutes
the environment and increases congestion. Relatively fewer proportions of the survey
respondents indicated that they have no reasonable alternative to driving most of the time
(32%) and that their schedule makes it hard or impossible for them to use public transit (28%).
These individuals are essentially car-dependent. Pro-environmental values were found to be
dominant in the sample population, with 80% or more showing that they cared about the
environment and that there is the need to control the rate of use of raw materials and reduce
travel-related energy use.
13
Table 2: Descriptive summary of participants’ responses to Likert Scale-based latent variables and their indicators
Latent concepts and indicators Response (%)
(Overall scale reliability: Cronbach’s Alpha α = 0.899) Strongly Disagree Neutral Agree Strongly
disagree Agree
Perceived benefits of ride-hailing (Instrumental/hedonic) (α = 0.889)
Ride-hailing would make traveling flexible 4 11 19 44 22
Ride-hailing would be reliable 5 10 23 41 21
Ride-hailing would be a fast option to travel 6 15 22 37 20
Ride-hailing would be a convenient way of travelling in my city 4 9 17 49 21
Ride-hailing would be suitable for non-work trips (e.g. shopping, a day-out, going to mosque/church, attending 5 9 12 48 26
parties, visiting family, friends etc.)
Ride-hailing would be suitable for work trips 6 8 17 46 23
Perceived benefits of ride-hailing (Environmental) (α = 704)
Ride-hailing could reduce congestion 10 13 16 41 20
Using ride-hailing could reduce environmental pollution associated with my travel 8 16 24 37 15
Ride-hailing could reduce the need for a personal vehicle 13 19 16 37 15
Environmental attitude and values (α = 0.894)
It makes me sad to see natural environment destroyed 2 1 8 32 57
I would like to reduce the consumption of energy and other resources while travelling 2 1 13 46 38
I am willing to spend a bit more to buy a product or use a service that is more environmentally friendly 2 2 13 44 39
In my daily life, I try to find ways to conserve energy 1 2 16 44 37
We need to control the rate at which raw materials are used to ensure that they last as long as possible 2 2 14 40 42
Attitude towards car-ownership/use (instrumental/hedonic) (α = 0.868)
The car provides privacy while travelling 2 1 8 33 56
The private car is a more comfortable mode of travel 2 1 11 31 55
I definitely want to own a car 3 2 12 28 55
The private car is a faster mode of travel 2 3 13 33 49
Using a private car is safer 1 3 13 35 48
Owning a car is a symbol of status in society 4 12 18 31 35
Attitude towards car-ownership/use (environmental) (α = 0.72)
Using the private car pollutes less (air and noise pollution) 10 22 28 28 12
Using a private car reduces congestion 21 24 19 23 13
Car-dependency (α = 0701)
Most of the time, I have no reasonable alternative to driving 13 22 33 22 10
My schedule makes it hard or impossible for me to use public transport 22 24 26 20 8
Innovation Openness (to new products and services) (α = 0855)
14
I like trying things that are new and different 3 5 19 48 25

I try new things when I see other people are also doing or using them 8 17 25 36 14
I try new things when I see someone in my family doing or using them 7 17 25 39 12
I usually try new things when I know my friends have done same 7 16 25 39 13
Information and Communication Technology (ICT)-Use (α = 0.757)
Having Wi-Fi internet everywhere I go is important to me 5 7 17 33 38
Getting around is easier than ever before with my smart phone 3 5 19 42 31
Other indicator items
Learning how to use new technologies is often frustrating 11 24 22 31 12
Ride-hailing is an innovative transport/mobility service 5 3 9 44 39
I am confident that I can use a ride-hailing service 6 7 13 45 29
Using ride-hailing would be easy for me 6 10 15 42 27
I would not feel safe using ride-hailing and travelling with complete strangers in a ride-hailing vehicle 10 16 18 28 28
15
5.2.2 Overview of ride-hailing use and user characteristics

Nearly three out of every five individuals in the survey sample (59%) indicated that they had
used ride-hailing (Fig 2a). Among this sub-sample of internet-based ride-hailers, 39%, 25%
and 36% had used the on-demand travel service more than a month, within a month and in
the last one week prior to filing the survey, respectively. Additional data on ride-hailing use
patterns over a period of one week preceding the date when respondents filled in the survey,
revealed that 36%, 30% and 14% had used ride-hailing service in one, two and three days of
the week, respectively. Another 8%, 4% and 2% had done so in four days, five days and six
days out of the week respectively, while the remaining 5% had done so every day of the week.
Based on this information, the respondents self-identified with one of three types of ride-
hailing users. About 11% of users indicated that they used ride-hailing for almost every trip.
About 53% alluded to being occasional users who do so a few times in a month while the
remaining 36% identified as infrequent users, doing so about once in a month.
A little over three quarters (76%) of ride-hailers in the sample did request a ride by themselves
for their most recent trip, using the App. About 21% of users had their reference trip
requested for them by another person, although they did indicate that they would have been
able to do so themselves if they had to. Despite indicating that they would not be able to use
the online App to request a ride by themselves, about 3% of the users were still able to use
the service, having been assisted by another person with the online booking.
Fig 2: Ride-hailing use: (a) in the total sample population (b) among four generational cohorts (c) among
income-groups (d) among car owners/users and non-car owners
16
Moreover, usage of ride-hailing services differed significantly across the four generational
cohorts (Fig 2b). The results show that a large proportion of users were in the Post-millennial
(18-24yrs)—25% and Millennial (25-39yrs)—66% cohorts. Among Generation X and Baby
Boomers, about 81% and 94%, respectively, had never used a ride-hailing service (Fig 2b). In
addition, across income-groups, the proportion of ride-hailing service users increase as
income increases. For example, among individuals with monthly earnings of ≤GH¢1,050, 69%
had used a ride-hailing service but among those with higher earnings (i.e. > GH¢1,050), at
least 81% used a ride-hailing service (Fig 2c). The proportion of ride-hailers was high between
both car owners/users and non-car owners. However, as shown in Fig 2d, the proportion of
individuals who own a car and used a ride-hailing service was 80%, while among non-car
owners and non-car owners who have access to car in their household, 51% and 66% had
used a ride-hailing service, respectively.
5.2.3 Structural equation model of the determinants of ride-hailing adoption and use
The SEM examine the extent to which the latent and socio-demographic variables captured
through the survey, explain individuals’ propensity to use ride-hailing services. Ride-hailing
adoption/use, the outcome variable in the SEM is dichotomous, coded as ‘used ride-hailing’
(1) and ‘never used ride-hailing’ (0). The latent variables and other Likert scale-based variables
are continuous variables as depicted on the respective elicitation scales in Table 2. Moreover,
the socio-demographic variables in the SEM are all dichotomously coded as follows: Age is
coded (1) for relatively younger group aged 18-39 years and (0) for relatively older groups,
40 years and older; Education is coded (0) for non-tertiary and (1) for tertiary, comprising
those currently enrolled in university and those who have completed tertiary education,
including advanced degrees. The other socio-economic variables are Income, coded (0) for
monthly earnings ≤GH¢1,050 and (1) for monthly earnings ≥GH¢1,051; Gender, coded Female
(1); Male (0); and Car ownership coded yes (1), no (0).
The SEM allows us to specify correlations among all the explanatory models first and to
subsequently regress them on the outcome variable (i.e. ride-hailing adoption/use). The full
path diagram of the SEM and a summary of the model’s estimates are presented in appendix
Fig A1 and Table A1, respectively. Several statistically significant associations were first
established in the measurement model including a positive relationship between perceived
instrumental/hedonic benefits and perceived environmental-related benefits of using ride-
hailing services (appendix Table A1). Individuals in our sample who considered ride-hailing to
be an innovative mobility solution also favoured car-ownership and use and were more likely
to agree that ride-hailing services would satisfy the instrumental and hedonic expectations
associated with their daily travel. Controlling for other factors, perception of innovativeness
of ride-hailing also correlated positively with perceived ease of use and individuals’ perceived
confidence in being able to request rides using the various online platforms. Indeed,
individuals who reported using ICT in their daily travel also found ride-hailing to be an
innovative mobility solution, and were more confident in their ability to use the service.
17
Moreover, the results show a small but statistically significant negative association between
the age-group of individuals and their confidence in and perception of ease of use of ride-
hailing services. This implies that older people (i.e. 40+ year-olds) believed it would not be
easy for them to use the available ride-hailing services. Education, however, had a small but
positive effect on perceptions of ease of use, suggesting that tertiary-educated individuals
were more confident in their ability to use ride-hailing services and generally agreed that it
would be easy doing so. Indeed, higher levels of educational attainments correlate positively
with factors that could make ride-hailing service use relatively easy, including internet access
and ICT use in travel; pro-car ownership attitudes; actual car-ownership rate in the sample;
and self-reported car-dependent lifestyles.
Furthermore, relatively younger people and individuals with pro-car ownership/use attitudes
in our sample were more likely to indicate that ride-hailing services would be beneficial from
the instrumental and hedonic perspectives. Quite paradoxically, the majority of the survey
respondents held pro-environmental attitudes and values, but such convictions correlated
positively with pro-car ownership/use convictions and perceived instrumental and hedonic
benefits of using ride-hailing services. This probably reveals a combination of cognitive
dissonance and a lack of non-motorized travel alternatives, which lead to environmental
convictions not always matching expectations that individuals would choose non-car-based
transport options, such as ride-hailing services.
After accounting for all the correlation and covariance effects among variables in the
measurement model, ten variables were found to have direct and statistically significant
explanatory effect on ride-hailing use in the SEM (Fig 3). The results show that perceived
environmental-related benefits of using ride-hailing had a negative explanatory effect on
their use. This means that although a significant proportion of the sample population agreed
that ride-hailing could reduce the need for a personal vehicle and thereby reduce congestion
and associated negative environmental impacts, this belief was not associated with the
likelihood of use of the existing ride-hailing services. In addition, this finding could possibly
reflect an underlying belief among individuals with pro-environmental attitudes that ride-
hailing is essentially car-based transport and that, like other motorized modes of transport,
they could still exert overall negative impacts on the environment.
Furthermore, general attitude towards car ownership and use, which was found to be
dominantly positive in the sample population, was associated with reduced use likelihood of
using ride-hailing services, but self-reported car-dependent lifestyles had the opposite effect.
This suggests that overall, while individuals in the survey sample might favour car ownership
and car-based transport, this does not necessarily lead to using ride-hailing services. Rather,
the propensity to use ride-hailing services was relatively higher amongst individuals who
indicated that they do not have an alternative to car-based transport and that their daily
schedule made it hard or impossible for them to use public transport.
18
The SEM results further revealed that perceived ease of use of ride-hailing services and
personal confidence in being able to do so was associated with increased likelihood of use of
ride-hailing services controlling for the other factors. Those who felt that it would be safe
using ride-hailing services in general and that it would as well be safe having to possibly share
a ride with strangers, were also found to be more likely to use the service.
Finally, among the socio-demographic factors, educational attainments, income-group,

gender and age-group had positive effects on the use of ride-hailing services. What these
results mean in the context of this study is that tertiary- educated individuals with relatively
higher incomes (≥GH¢1,051) were more likely to use ride-hailing services. Moreover, being
female was also associated with increased likelihood of use of ride-hailing services.
Controlling for the other factors, relatively younger people (i.e. 18-39 year-olds) were far
more likely to use ride-hailing services.
Fig 3: A simplified path diagram of the SEM showing standardised co-efficient of only the factors that have
statistically significant direct effect on ride-hailing adoption and use.
Model fit indices: χ2 = 2601.315, df = 671, Normed-χ2 =3.876, p < 0.001; RMSEA = 0.049; NFI = 0.910; CFI =
0.921; IFI = 0.922; TLI = 0.904, *** P-value < 0.01, **P-value < 0.05
The detailed path diagram and table summary of the model’s estimates are provided in appendix Fig A1 and
Table A1.
5.3 Characteristics of on-demand ride-hailing trips

In addition to the factors influencing ride-hailing use explored in the previous section, this
section presents detailed analysis of the characteristics of ride-hailing journeys, by focusing
on users’ most recent single reference trip for which they used a ride-hailing service.
19
5.3.1 Purpose of trips and reasons for ride-hailing

Ride-hailing service users indicated the main purpose of their most recent trip for which they
used a ride-hailing vehicle and the reasons for doing so4. The results show that a little over
two-fifth (41%) of individuals’ most recent ride-hailing trips involved various forms of work
and school-related (mainly university) trips—the core journeys that many individuals
undertake daily (Fig 4a). About 50% of the reference ride-hailing trips could be classified as
‘special occasion trips’, which were for purposes including, social functions, such as weddings
and parties; visiting friends and family; and church or mosque attendance (Fig 4a). The
purpose of 9% of the trips was undisclosed by the respondents and are, therefore, unknown.
In addition to ride-hailing use among the adult respondents, the survey also explored whether
those with children of school age, had ever requested ride-hailing services for their children’s
school journeys. The data showed that among the respondents with children (n = 418), only
a few of them (12%) had booked a ride-hailing service, at least once, for their children’s school
trip. For the five-day period preceding the survey, about 57% and 21% of the children of this
sub-sample population had used ride-hailing services for travel purposes once and twice in a
week, respectively. The remaining 18% and 4% had used this mode to travel to school three
and five times over the five-day period, respectively. All the school trips involving ride-hailing
vehicles were return journeys.
The reasons stated by the respondents for using ride-hailing for their own reference trips and
that of parents who booked rides for their children’s school trips are summarised in Figs 4b
and 4c, respectively. The results showed that more than three quarters of all the reasons
given by the respondents for using ride-hailing for the reference trips, relates to instrumental
utilitarian and hedonic benefits, including ride-hailing being convenient, cheaper, reliable,
comfortable and quicker travel option, compared to other available public transport modes.
Other reasons given were the fact that the ride-hailing services afford privacy, which is valued
by those indicating it as a reason (3.5%) and difficulty getting public transport (3.6%) for the
trip in question. Moreover, some of the users indicated that using ride-hailing is classy (4.1%),
and hip and trendy (0.8%), and that TNC drivers provided mobility services professionally. The
unavailability of private cars, due to having sent them for maintenance and repairs and
weather-related reasons, including the weather being too hot to walk or heavy downpours at
the time of the reference trip were also cited as reasons for ride-hailing (Fig 4b).
In relation to children’s school journeys, one of the common reasons cited by the respondents
was that, even though they had a car, they were usually busy to drive their children to school.
Others either left home for work too early in the morning or returned home so late in the
evening that it would be impossible to pick their children from school. Besides these, most of
the reasons given also relate to hedonic benefits, such as convenience, reliability and
flexibility of using ride-hailing services (Fig 4c).
4
It was clarified to the respondents that the reference trip was the single trip starting where the Uber/Bolt
driver picked them and ended where they alighted and paid for the trip.
20
In order to appreciate some of the underlying reasons given by the respondents for using ride-
hailing for reference trips, it is important that we situate them in the wider urban transport
and mobility contexts of the case study metropolis. Firstly, in other contexts such as Europe
and North America, similar ICT-enabled mobility services existed prior to the presence of TNCs
(i.e. Uber, Lyft etc.). For example, individuals were able to call a central fleet management to
request a taxi or even do so using a smartphone app. In the Ghana context, however, these
earliest forms of city-wide on-demand mobility services had never existed prior to the entry
of TNCs, including Uber and Taxify/Bolt. Instead, conventional taxis could only be taken at
designated stations or by hailing them from the street. Thus, for most of the users, ride-hailing
is very revolutionary, making it possible, for the first time, to access such convenient and
flexible door-to-door mobility services. In addition, the view that ride-hailing is a faster option
could be related to the fact that on-demand ride-hailing significantly reduces vehicle access
and waiting time compared to station-based taxis, thereby reducing overall travel time.
Additional supporting evidence from the survey regarding waiting times in ride-hailing
services are provided later in section 5.3.2
21
Fig 4: (a) Ride-hailing trip purpose (b) reasons for using a ride-hailing service for adult respondents’ reference trip; (c) reasons for using a ride-hailing service for school
journeys by children of respondents
22
Fig 5: Travel costs associated with reference ride-hailing trips
23
Another important contextual factor that helps place the above findings in perspective is the
out-of-pocket costs associated with using ride-hailing services on the one hand and using
conventional taxi and/or public transport (Trotro) on the other hand. As shown in Fig 5, out-
of-pocket costs incurred vary for ride-hailing journeys within the given travel times. On the
average, users paid GH¢8.00 ($1.45) for journeys lasting up to 10 minutes and GH¢21($3.82)
for ride-hailing journey times between 25 and 30 minutes. These ride-hailing costs are
compared with what individuals would have incurred if they were to use public transport
(Trotro) for selected actual reference trips reported in the survey. For example, for a typical
ride-hailing trip that lasted about 10 minutes and costs about GH¢8.00 ($1.45), one would
have incurred only GH¢2 ($0.35) if they were to use the existing public transport (Trotro)5.
Similarly, for another 25-30 minutes ride-hailing journey that costs GH¢21($3.82) on the
average, an individual would have incurred about GH¢4($0.72) for the same journey, using
Trotro. What the above comparisons show is that ride-hailing’s out-of-pocket costs are
significantly higher than what public transport users who substituted for ride-hailing, for
example, would have incurred. Despite this, some of the respondents used ride-hailing
instead of public transport for the non-monetary benefits already outlined.
Furthermore, even for private car and conventional taxi users, the evidence suggests that
using a ride-hailing service could as well be a relatively cheaper alternative. For example, for
a typical 25-30 minutes lone travel in a conventional taxi that is hailed on the street (known
locally as ‘dropping’ or ‘hire’), one would have incurred an out-of-pocket cost of about GH¢80
($13.16)6 compared to using one of the available internet-based riding-hailing that costs
around GH¢21 ($3.82) and provides private, on-demand and convenient mobility. This
explains why one of the reasons for ride-hailing cited by the respondents was that it was a
cheaper option. Overall, most of the users indicated that they received value for the fares
charged and that they were satisfied (56%) or very satisfied (20%) with the overall quality of
ride-hailing services. Whereas 20% of users were ambivalent regarding overall satisfaction,
only 3% and 1% indicated that they were either ‘not satisfied’ or not ‘satisfied-at-all’ with the
ride-hailing service they paid for.
5.3.2 Ride-hailing travel characteristics—journey start times, travel time, vehicle

occupancy and trip flows
Results of the analysis on journey start times for the reference ride-hailing trips show that
about two fifths (40%) of journeys occurred in the morning while the remaining 31% and 29%
occurred in the afternoon and evening times, respectively. Out of the total reference trips,
6%, 17% and 17% occurred in the hours between 5-7am, 7-9am and 9 -11am, respectively. A
further 16% and 15% occurred in the afternoon between the hours of 12-2pm and 2-5pm,
respectively. Thus, it can be inferred that for morning ride-hailing trips, more than half (59%)
5 The GH¢2($0.35) out-of-pocket travel cost estimate for public transport (Trotro) journey is based on an actual reported
journey from Legon to Haatso in Accra, covering approximately 4.9km. While that of the GH¢4($0.72) out-of-pocket travel
cost is based on an actual reported journey from Madina to Achimota in Accra, covering approximately 16.3km.
6 The GH¢80 ($13.6) out-of-pocket travel costs are estimates for a conventional taxi hire based on an actual reported journey
from Weija to Ministries in Accra, covering approximately 20km.
24
would have occurred during the morning peak travel period while 41% occurred in the late
morning off-peak travel period. Similarly, for afternoon trips, about 53% occurred in the
afternoon peak travel period. More than three fifths (62%) of the evening time reference trips
occurred in the evening peak period from 5pm (Fig 6a).
The data on reported waiting times after requesting a ride sheds light on why the majority of
the users cited convenience, flexibility and overall travel time as some of the main reasons
for choosing a ride-hailing service for their reference trips. This is because, as shown in Figure
6b, the wait times between requesting a ride and being picked up by drivers was quite low:
less than 5 minutes for 65% of respondents, and between 5 and 10 minutes for 42% of
respondents. The waiting time for an additional 17% of all online vehicle requests were
reported to have been met between 10 and 15 minutes by the service providers (Fig 6b).
Regarding reported journey times, the data show that individuals’ reference ride-hailing trips
would have covered relatively shorter distances. Indeed, 78% of the surveyed ride-hailing trips
took between 10 and 30 minutes (only considering in-vehicle travel times)—Fig 6c. Similar to the
adult respondents’ own ride-hailing trips, the majority of school trips that involved ride-
hailing use covered relatively shorter travel times, with 59% lasting less than 30 minutes. The
remaining 38% and 3% of school journeys involving ride-hailing use lasted between 30
minutes and 1 hour, and 1-2hrs, respectively.
The survey also explored vehicle occupancy for the reference ride-hailing trips, based on the
reported number of passengers in the vehicle at the individuals’ reference trips’ origins and
destinations7. Regarding occupancy at origin, 87% of the respondents indicated that there
were zero passengers in the vehicle when it arrived to pick them (Fig 5d). Similarly, at the
destination end of the reference trip, 85% indicated that they were the only passenger in the
taxi (Fig 5e). These findings from the data suggest that a significant share of ride-hailing trips
involved single occupancy of the vehicle. This is in contrast with traditional station-based taxi
services in which lone travel is uncommon, but instead, taxis tend to load fully for a given
journey. The predominance of single occupancy travel is a result of the fact that ride-sharing
services were not offered by TNCs as of the time of the survey. Even so, a question in the
survey that explored the respondents’ safety perceptions regarding ride-sharing revealed
some insight into a possible barrier to sharing, if such services were to become available.
Among users of this new transport service, 58% indicated that they would not feel safe while
ride-sharing with complete strangers. Only 24% of the respondents indicated that they would
feel safe with ride-sharing services, while the remaining 18% were ambivalent about ride-
sharing. This is notwithstanding the fact that most of the respondents, especially those who
use conventional taxis, are familiar with sharing a ride with complete strangers.
7Individuals were asked to indicate the number of passengers in the vehicle at the reference trip’s origin and destination,
excluding themselves and the driver. Zero passengers as shown in Figs 5d and 5e means that except the respondent and
the driver, there were no other persons in the vehicle.
25
Fig 6: Ride-hailing travel characteristics showing reported (a) journey start times (b) waiting time after vehicle requests (c) total travel times (d) vehicle occupancy at origin
of reference trip and (e) vehicle occupancy at destination of reference trip
26
Fig 7: Reference ride-hailing trips distributions among three broad zones in (a) Kumasi and (b) Accra
Finally, ride-hailing trip flows obtained from the survey were mapped, using data on
individuals’ reported localities of origin and destination for their reference trips. Fig 7 presents
a simplified illustration of the flows, based on three broad urban zones, namely the inner-city,
inner-suburban and outer-suburban areas. Overall, the reference ride-hailing trips were
generated and distributed across all three zones, although a significant proportion of trips
occurred in the suburban zones. For example, 70% and 64% of ride-hailing trips originated
from an outer-suburban locality in the Kumasi and Accra metropolis respectively. In addition,
while some of the reference trips in the survey data cut across the different zones, the origin-
destination flows show that a significant proportion of the reference trips have the same
zones of origins and destinations. For example, 63% and 54% of the trips that originated in
the outer-suburban zone in both metropolitan areas had destinations in the same zone. For
Accra, about 43% of all the reference ride-hailing trips that started in the inner-city zone
ended in the same zone. These findings support the earlier finding that most ride-hailing trips
tend to cover relatively shorter travel times and hence shorter travel distances.
The observed ride-hailing trip flows could stem from the wider public transport accessibility
challenges in the two case study areas. Typical suburban localities represented in our ride-
hailing trip flow data included the public University Campuses in each of the study areas, and
their relatively affluent catchment neighbourhoods. In these two metropolitan areas, as is
also typical in urban areas in Ghana, the Trotro public transport services are organized around
stations, which tend to be located within high population density areas in the metropolitan
core. This means that most suburban areas are not adequately covered by conventional public
transport services. Conventional public transport does not also run on a timetable, which
means that commuters face uncertainties around reliability when accessing public transport,
especially for urgent travel needs. For these reasons, suburban localities tend to be car-
27
dependent, with most commuters relying on the conventional taxi and/or the private car. In
the absence of car-based modes of transport, accessing the station-based Trotro public
transport or doing so by hailing one by the roadside, could involve considerable walking
distance and access time. Indeed, as the data has already shown, one of the reasons given by
the respondents for using ride-hailing services is the fact that they could not access
conventional public transport for their reference trips. Thus, by offering on-demand door-to-
door mobility in these suburban locations, ride-hailing becomes a convenient and flexible
alternative mode of transport, enabling users to overcome the existing public transport
accessibility challenges.
5.3.3 Mode substitution effect of ride-hailing and multi-modality

The final aspect of the analysis explores mode substitution effects of the presence of ride-
hailing. First, users of the service indicated one of the existing conventional modes that they
typically would have used for their reference trips before ride-hailing services became
available. The data shows that ride-hailing replaced conventional taxis, public transport
(minibus/Trotro) and bus for 51%, 36% and 1% of the reference trips of adult respondents of
the survey, respectively. Moreover, about 10% of the reference trips would have been
completed using private cars in the absence of ride-hailing services. A further 1% of users
indicated that they would have walked the entire way for the trip in question, if they did not
have the option of requesting a ride-hailing service online (Fig 8a).
Similar findings of mode substitution effect were also evident from the data on children’s
school trips that involved ride-hailing use. For the reference school trips, ride-hailing replaced
43%, 36% and 9% of trips that would have involved the use of a private car, a conventional
taxi and public transport (Trotro), respectively. The remaining 12% of school trips involving a
ride-hailing would previously have been completed using the school bus (6%) or walking as
the main mode of transport (6%). In all cases, the survey respondents indicated using the ride-
hailing to complete the full reference trip. This means that individuals used ride-hailing to
meet the single reference door-to-door trip, instead of as first/last mile feeders to other
modes of transport within a trip chain.
In addition to the substitution analysis presented above, individuals also indicated if they used
any other travel mode(s) on the day of their reference ride-hailing trips. Fig 8b shows the
number of times each of the other available modes was selected, while Fig 8c provides a
breakdown in terms of mode combinations for the full day’s travel choices. The results show
that 92% of those who used a ride-hailing service also used at least one of the conventional
modes of transport for other trips they undertook on the same day as their reference ride-
hailing trip. The majority of ride-hailing users used public transport (Trotro/Mini-bus) (33%),
traditional taxis (23%) and private car (23%) for other trip purposes on the same day as their
reference ride-hailing trips (Fig 8b). Furthermore, ride-hailers also used public transport
(Trotro/Minibus) and walking (28%), and public transport (Trotro/Minibus), traditional taxi,
walking and motorcycle (25%) on the same day as their reference trip captured in the survey
(Fig 8c).
28
The above findings from the survey show that from the point of view of a full day’s travel
choices, users of ride-hailing services also use other available modes for other journeys.
However, as explained previously, each trip was completed using a single mode of transport,
implying that for the trips they replaced, ride-hailing took away passengers from existing
conventional public transport alternatives (i.e. taxi and Trotro). In addition, this new form of
transport was utilised as independent alternatives and did not necessarily feed passengers to
the public transport alternatives.
29
Fig 8: (a) modes substituted with ride-hailing (b) other modes used by ride-hailers on the day of reference trip (c) breakdown of travel mode combinations used in
addition to ride-hailing on the day of individuals’ reference trip
30
6 Discussion and implications of findings

This paper has provided detailed empirical insights into ride-hailing adoption and use
patterns. The study focused on a developing country context in Sub-Saharan Africa, where
new on-demand ride-hailing services are diffusing rapidly, but an understanding of the factors
driving adoption, as well as use patterns and emerging travel behaviour impacts are limited.
A theory-grounded conceptual model was mobilised to empirically model the behavioural
factors behind ride-hailing adoption, using a structural equation modelling approach. To
deepen our understanding beyond the adoption factors, important aspects of reference ride-
hailing journeys, including trip purpose, journey times and trip flows, as well as mode
substitution effects associated with the diffusion of ride-hailing services have been examined.
The key findings and their implications are discussed as follows, weaving together the results
of the SEM on ride-hailing adoption and that of the descriptive statistics on travel
characteristics.
The SEM of ride-hailing services adoption revealed that users and non-users are distinguished
by socio-demographic characteristics (i.e. education, age, income and gender); perception of
the benefits and ease of use of ride-hailing services; perception of the safety risks associated
with using ride-hailing; car ownership/use attitudes; and self-reported car-dependent
lifestyles. Consistent with the findings of previous studies (e.g. Ackaah et al., 2020; Grahn et
al., 2019; Lavieri and Bhat 2019; Schaller, 2018; Lewis and MacKenzie, 2017), the empirical
analysis presented in this paper suggests that being young, highly educated with relatively
higher earnings are associated with higher propensities to use ride-hailing services. As our
survey data has shown, these demographic groups are more open to and familiar with new
products and technologies as evidenced by their ownership of smart phones, ability to afford
internet access and use of ICT to aid their daily travels. Furthermore, as the analysis revealed,
these relatively tech-savvy individuals in the population also tend to be more confident to be
able to request rides using the relevant digital platforms and agreed that they would find it
easy to use ride-hailing services.
In our case study areas, ride-hailing services have become available in the context of an
overall perceived inferior quality of existing public transport services (see e.g. Acheampong,
2020; Esson et al., 2016). As the survey data showed, about 37% of all the reference ride-
hailing trips would have previously been completed using public transport. Thus, it seems
young adults and professionals are taking advantage of the availability of ride-hailing services
to meet mobility service quality gaps that they hitherto experienced. This is notwithstanding
the fact that doing so costs significantly higher than using public transit (Trotro). Indeed,
dominant among the reasons given for using ride-hailing services were instrumental
utilitarian and hedonic advantages, including ride-hailing being convenient, reliable,
comfortable, classy and quicker travel option, as well as the professionalism of drivers. These
factors are critical indicators of quality in any mobility service and influence user satisfaction.
31
Consequently, the majority of users also indicated being satisfied with the level of service they
receive in using the available ride-hailing services. Ride-hailing replacing public transport
could have negative consequences, including increased motorization, congestion and
environmental pollution. It is clear from our findings that the innovation introduced by TNCs
certainly brings to the user superior mobility service benefits that compared to the existing
conventional taxi and public transport service. Thus, to avert the potential negative
consequences of car-dependence resulting from increased adoption and use of ride-hailing
services, the quality of the existing public transport would have to improve significantly,
especially in suburban localities where access to public transit is a challenge. This would
require interventions in planning and investments that go beyond the adoption of ICTs per
se. With these fundamental interventions that brings about schedule and reliable public
transport services, passengers could leverage ICT to access public transport as they do with
ride-hailing services.
The SEM analysis further revealed that individuals with car-dependent lifestyles were more
likely to use ride-hailing services. This finding, together with the findings regarding mode
substitution effects of ride-hailing helps to gain a better understanding of why this is so and
the implications. As our data showed, about 10% and 51% of individuals’ reference ride-
hailing trips would have been undertaken in the past, using the private car and conventional
taxi, respectively. These two conventional modes are also essentially car-based forms of
transport that are used by individuals, mostly suburbanites, who reported car-dependent
lifestyles. Thus, together, these findings suggest that ride-hailing services appear to promote
and entrench car-based transport among individuals who already have a car or those who do
not, but feel the need to use the car due to their daily schedules or the lack of alternatives to
car-based transport. Furthermore, it is likely that even when people substitute ride-hailing for
private car use, for example, overall levels of motorization do not decrease as a result. This is
because, as our findings show, the majority of ride-hailing trips involve single occupancy of
the vehicle. In addition, individuals switching from the conventional taxi to ride-hailing could
as well lead to an increase in motorization. This is because, in the Ghana context, conventional
station-based taxis tend to take more passengers, usually up to four before they leave their
stations, while the new on-demand ride-hailing typically transport a single passenger for a
given trip. These findings imply that mode substitution behaviours triggered by the availability
of ride-hailing service do not necessarily reduce car use and any possible negative societal
effects that might result from car-based transport. Instead, they could be contributing to
congestion that typifies urban mobility in Ghana’s major cities.
Previous studies have found that ride-hailing is commonly used for special occasion, social
and recreational trips (Grahn et al., 2019; Hampshire et al., 2018; Rayle et al., 2016). While in
this study about half of the reference trips using ride-hailing fall into these categories, the
data also revealed that ride-hailing for work and school travel purposes was also high (i.e.
41% of all reference trips). Together, these findings suggest that compared to other contexts
32
where similar studies have been undertaken, ride-hailing use for commuting purposes by
adult respondents in urban areas in Ghana might be significantly higher. The high incidence
of ride-hailing use for commuting purposes in this study could be because of the fact that car-
ownership is quite low (i.e. 15% among the survey sample) and public transit, as already
highlighted, is of relatively poorer quality.
The results further show that a significant proportion of ride-hailing journeys covered
relatively shorter travel times (i.e. below 30 minutes) and occurred mostly in inner and outer-
suburban localities. Indeed, previous studies have shown that intra-urban commuting trips
tend to cover relatively shorter distances in Kumasi (i.e. 4.5kms) and Accra (5-9kms), the two
metropolitan areas represented in this study (see e.g. Acheampong, 2020; Agyemang 2017).
Our findings, however, seem to contradict that of Clewlow and Mishra (2017) who found that
in the American context, ride-hailing is used frequently by urban dwellers and less so by those
in the suburbs. By inferring from the reported travel times and comparing them with peak
and off-peak traffic trends (see Adarkwa and Poku-Boansi, 2011; Department of Urban Road,
2004), it is could be concluded that ride-hailing trips broadly typify overall mobility trends in
the metropolitan areas. Analysis of the journey start times of the reference ride-hailing trips
also revealed that they typify the overall peak and off-peak trends experienced in the two
metropolitan areas represented in this study. Thus, similar to travel demand associated with
other modes, ride-hailing trips follow space-time constraints, such as authority constraints
imposed, for example, by work or school reporting and closing times. Consequently, the use
of ride-hailing services could as well, contribute to overall emergent traffic conditions, such
as congestion experienced during peak travel periods in the two metropolitan areas.
Moreover, major service providers, including Uber and Bolt do not offer ride-sharing options
in the case study areas. This partly explains the high proportion of lone travel that typifies
ride-hailing trips in our survey. At the same time, we found that a significant proportion of
the respondents indicated that they would not feel safe having to share a ride with strangers.
It is rather surprising that people are willing to enter conventional station-based taxis that
have strangers in them, but they are reluctant to do same with ride-hailing services that offer
shared rides. Our data does not allow us to explain why this is so and this will require further
investigation as we highlight in the final section of this paper. That said, the perceived safety
risks around ride-hailing implies that even when ride-sharing options become available, they
could be a barrier to increasing vehicle occupancy levels in ride-hailing services. As lone
occupancy is undesirable from an environmental and sustainability standpoint, it would be
crucial for urban transport regulators to ensure that TNCs provide ride-sharing options, as is
the case with conventional station-based taxi services. In addition, the TNCs themselves (e.g.
Uber and Bolt) would have to improve their safety measures by fixing security lapses in their
platforms for example, as a way of building trust and encouraging more people to share.
33
Finally, by examining a full day’s travel mode choice, our data revealed that individuals use
other available modes of transport in addition to ride-hailing services. However, our data also
revealed that ride-hailing services do not necessarily link with other available modes as
feeders to address first and last mile passenger mobility solutions for example. Instead, they
form a part of a transport system configuration in urban Ghana, whereby multiple modes
exist, but the individual modes remain standalone options that do not functionally integrate
with each other. This means that the potential of ride-hailing and new ride-sharing services
to complement other modes, such as public transport, thereby increasing public transit use,
is currently not being realised. Given that integration with public transit could help maximize
the overall societal benefits of these new forms of mobility services, it might be worth
exploring possible ways of realising this imperative. One possible strategy could be using ride-
hailing services as feeders to public transit, especially in suburban areas, where first/last mile
access to public transport remains a challenge.
7 Conclusion, limitations and further work

This paper has contributed to our understanding of ride-hailing adoption factors, use patterns
and mode substitution effects from a developing country context. The survey-based evidence
shows that this new form of transport provides essential mobility services and help bridge
existing transport service quality gaps. However, low-vehicle occupancy and mode
substitution behaviours that result in ride-hailing replacing car-based transport (i.e. private
car and conventional taxis) while taking away passengers from existing public transport
services, suggest that their presence could possibly generate negative unintended effects.
Implementing ride-sharing services, overcoming safety perceptions that could act as barriers
to internet-based ride-sharing schemes, and integrating this new form of mobility services
into other modes and services could help optimise overall benefits from the urban
transportation system and capture public value from the on-going digital platform mobility
transitions.
Notwithstanding the detailed empirical insight provided in this paper, there are limitations
worth highlighting for further research. The empirical evidence provided in this paper is based
on a cross-sectional survey data that is not weighted to the total population of the case study
areas. Thus, while the sample is large enough and the resulting data does enable a snapshot
understanding of the topic, it is not necessarily representative of the whole population. This
also means that the current data is limited in helping us to understand some of the wider
impacts of the presence of ride-hailing services on the transportation system and society. In
addition to large-scale surveys, TNCs (e.g. Uber, Bolt) making data generated on their
platforms readily available could help address this limitation in future studies. Further
research could help generate an understanding of the travel behaviour changes being
triggered by ride-hailing services, by focusing on individuals’ activity-travel choices over time.
This will help generate a better understanding of both the mode substitution and
complementarity effects associated with ride-hailing services. Ultimately, travel demand
modelling and travel management strategies would benefit from these detailed empirical
34
insights. The current study and some previous work have focused mainly on use patterns of
ride-hailing services among individual adult respondents, but as this study has shown, other
members of the household, including children could be using this new form of mobility. In
particular, the evidence provided in this study that ride-hailing service also are being used for
children’s school journeys, could be worth exploring further in different contexts. Finally, in
the context of this study, when individuals use conventional station-based taxis, they do so
sharing a ride with other strangers. Yet, they are uneasy about ride-sharing with the new form
of ICT-mediated mobility service because of perceived safety risks. Future research could
examine safety perceptions and experiences around ride-hailing, as this could be crucial to
successfully implementing ride-sharing solutions.
Appendix A: Path diagram and table summary of the results of the SEM of
ride-hailing adoption factors
Fig A1: Path diagram of SEM of the determinants of ride-hailing mobility service use
Model fit indices : χ2 = 2601.315, df = 671, Normed-χ2 =3.876, p < 0.001; RMSEA = 0.049; NFI = 0.910; CFI =
0.921; IFI = 0.922; TLI = 0.904.
35
Table A1: Correlations, covariance and regression weights of relationships among variables in SEM of ride-
hailing mobility service use
A. Direct Effects ß b S.E. C.R. P-Value
Ride-hailing service use <-- Perceived ride-hailing benefits -0.041 -0.071 0.019 -2.158 0.031
(environmental)
Ride-hailing service use <-- Innovation openness (new products -0.009 -0.017 0.014 -0.614 0.539
and services)
Ride-hailing service use <-- Attitude towards car ownership/use -0.083 -0.100 0.030 -2.725 0.006
(hedonic)
Ride-hailing service use <-- Attitude towards car ownership/use -0.025 -0.051 0.023 -1.075 0.282
(environmental)
Ride-hailing service use <-- Car-dependent lifestyle 0.068 0.130 0.026 2.605 0.009
Ride-hailing service use <-- Perceived ride-hailing benefits -0.011 -0.016 0.028 -0.389 0.697
(instrumental/hedonic)
Ride-hailing service use <-- Age-group 0.244 0.223 0.034 7.112 ***
Ride-hailing service use <-- Education 0.321 0.286 0.042 7.673 ***
Ride-hailing service use <-- Environmental attitude and values 0.013 0.018 0.027 0.467 0.640
Ride-hailing service use <-- ICT use in travel -0.005 -0.010 0.019 -0.269 0.788
Ride-hailing service use <-- New technology use-frustrating -0.004 -0.011 0.011 -0.407 0.684
Ride-hailing service use <-- Ease of use easy of ride-hailing 0.051 0.120 0.017 3.024 0.002
Ride-hailing service use <-- Perceived control/confidence to use 0.032 0.076 0.017 1.875 0.041
ride-hailing
Ride-hailing service use <-- Ride-hailing is innovative -0.017 -0.034 0.017 -1.016 0.310
Ride-hailing service use <-- Gender 0.071 0.074 0.024 2.946 0.003
Ride-hailing service use <-- Income-group 0.078 0.074 0.028 2.807 0.005
Ride-hailing service use <-- Car ownership 0.059 0.044 0.035 1.663 0.096
Ride-hailing service use <-- Perceived safety risks of ride-hailing 0.022 0.059 0.009 2.483 0.013
B. Correlations and covariance Cov Corr S.E. C.R. P-Value
Perceived ride-hailing benefits <--> Perceived ride-hailing benefits 0.262 0.463 0.024 10.725 ***
(instrumental/hedonic) (environmental)
Perceived ride-hailing benefits <--> Innovation openness (new products 0.080 0.124 0.018 4.511 ***
(instrumental/hedonic) and services)
Perceived ride-hailing benefits <--> New technology use-frustrating 0.112 0.139 0.019 5.796 ***
Perceived ride-hailing benefits <--> Perceived ease of use easy of ride- 0.257 0.339 0.023 11.316 ***
(instrumental/hedonic) hailing
Perceived ride-hailing benefits <--> Ride-hailing is innovative 0.344 0.525 0.023 14.763 ***
Perceived ride-hailing benefits <--> Perceived control/confidence to use 0.217 0.286 0.021 10.128 ***
(instrumental/hedonic) ride-hailing
Perceived ride-hailing benefits <--> Attitude towards car ownership/use 0.115 0.293 0.014 8.472 ***
(instrumental/hedonic) (hedonic)
Perceived ride-hailing benefits <--> Car-dependent lifestyle -0.107 -0.173 0.020 -5.254 ***
Perceived ride-hailing benefits <--> Environmental attitude and values 0.136 0.297 0.015 9.276 ***
(instrumental/hedonic) (environmental)
Perceived ride-hailing benefits <--> Education -0.072 -0.250 0.007 -10.141 ***
Perceived ride-hailing benefits <--> Age -0.067 -0.226 0.008 -7.878 ***
Perceived ride-hailing benefits <--> Gender 0.027 0.080 0.009 3.138 0.002
Perceived ride-hailing benefits <--> Gender -0.033 -0.079 0.012 -2.773 0.006
(environmental)
Perceived ride-hailing benefits <--> Car-dependent lifestyle 0.109 0.142 0.027 3.983 ***
(environmental)
(environmental) (environmental)
36
Perceived ride-hailing benefits <--> Environmental attitude and values 0.091 0.160 0.018 5.073 ***
(environmental)
Perceived ride-hailing benefits <--> Innovation openness (new products 0.155 0.194 0.025 6.290 ***
(environmental) and services)
Perceived ride-hailing benefits <--> Ride-hailing is innovative 0.133 0.163 0.021 6.339 ***
(environmental)
Perceived ride-hailing benefits <--> Perceived ease of use easy of ride- 0.104 0.111 0.018 5.837 ***
(environmental) hailing
Perceived ride-hailing benefits <--> Age -0.043 -0.116 0.009 -4.822 ***
(environmental)
Environmental attitude and values <--> Car-dependent lifestyle -0.032 -0.052 0.017 -1.903 0.047
Environmental attitude and values <--> ICT use in travel 0.207 0.345 0.020 10.405 ***
Environmental attitude and values <--> Innovation openness (new products 0.095 0.146 0.019 4.970 ***
and services)
Environmental attitude and values <--> Age -0.023 -0.079 0.007 -3.349 ***
Environmental attitude and values <--> Ride-hailing is innovative 0.160 0.243 0.017 9.394 ***
Environmental attitude and values <--> Attitude towards car ownership/use 0.247 0.627 0.020 12.375 ***
(hedonic)
Attitude towards car <--> Attitude towards car ownership/use 0.102 0.179 0.017 6.044 ***
ownership/use (hedonic) (environmental)
Attitude towards car <--> Ride-hailing is innovative 0.129 0.229 0.015 8.570 ***
ownership/use (hedonic)
Attitude towards car <--> Innovation openness (new products 0.084 0.153 0.016 5.164 ***
ownership/use (hedonic) and services)
Attitude towards car <--> Age -0.035 -0.137 0.007 -4.937 ***
Attitude towards car <--> Education -0.035 -0.140 0.006 -5.945 ***
Attitude towards car <--> ICT use in travel 0.103 0.202 0.016 6.274 ***
Perceived ride-hailing benefits <--> Attitude towards car ownership/use 0.042 0.086 0.015 2.845 0.004
(environmental) (hedonic)
Attitude towards car <--> Perceived safety risks of ride-hailing 0.113 0.147 0.020 5.768 ***
Attitude towards car <--> New technology use-frustrating 0.281 0.239 0.038 7.362 ***
ownership/use (environmental)
Attitude towards car <--> Innovation openness (new products 0.262 0.280 0.035 7.581 ***
ownership/use (environmental) and services)
Attitude towards car <--> Education -0.106 -0.252 0.014 -7.849 ***
Attitude towards car <--> Age -0.107 -0.248 0.014 -7.402 ***
Attitude towards car <--> Car-dependent lifestyle 0.451 0.502 0.040 11.189 ***
Car-dependent lifestyle <--> Ride-hailing is innovative -0.111 -0.125 0.021 -5.214 ***
Car-dependent lifestyle <--> Education 0.097 0.249 0.013 7.719 ***
Car-dependent lifestyle <--> Gender -0.107 -0.233 0.015 -7.178 ***
Car-dependent lifestyle <--> Income-group 0.052 0.125 0.013 4.104 ***
Car-dependent lifestyle <--> Innovation openness (new products 0.143 0.165 0.029 4.951 ***
and services)
Car-dependent lifestyle <--> Car ownership 0.068 0.206 0.010 6.527 ***
Car-dependent lifestyle <--> ICT use in travel 0.211 0.262 0.028 7.521 ***
Car-dependent lifestyle <--> Age 0.049 0.122 0.013 3.829 ***
Innovation openness (new <--> New technology use-frustrating 0.350 0.308 0.034 10.184 ***
products and services)
Innovation openness (new <--> Education -0.066 -0.163 0.011 -5.808 ***
Innovation openness (new <--> Age -0.049 -0.118 0.012 -4.082 ***
Innovation openness (new <--> ICT use in travel 0.166 0.198 0.027 6.244 ***
New technology use-frustrating <--> Education -0.169 -0.331 0.014 -11.925 ***
New technology use-frustrating <--> Gender 0.067 0.112 0.016 4.274 ***
37
New technology use-frustrating <--> Income-group -0.068 -0.125 0.014 -5.065 ***
New technology use-frustrating <--> Age -0.157 -0.300 0.015 -10.556 ***
Ride-hailing is innovative <--> Perceived control/confidence to use 0.638 0.585 0.035 18.199 ***
ride-hailing
Ride-hailing is innovative <--> Perceived ease of use easy of ride- 0.586 0.539 0.034 17.323 ***
hailing
ICT use in travel <--> Perceived control/confidence to use 0.210 0.212 0.028 7.468 ***
ride-hailing
ICT use in travel <--> Ride-hailing is innovative 0.114 0.133 0.022 5.233 ***
ICT use in travel <--> Car ownership 0.024 0.076 0.008 2.999 0.003
ICT use in travel <--> Gender -0.094 -0.213 0.013 -7.234 ***
ICT use in travel <--> Age 0.101 0.262 0.012 8.710 ***
ICT use in travel <--> Perceived ease of use easy of ride- 0.236 0.239 0.028 8.362 ***
hailing
ICT use in travel <--> Education 0.126 0.334 0.011 11.647 ***
Perceived ease of use easy of ride- <--> Age 0.057 0.117 0.012 4.637 ***
hailing
Perceived ease of use easy of ride- <--> Education 0.062 0.129 0.010 6.134 ***
hailing
Perceived control/confidence to <--> Education 0.038 0.079 0.010 3.962 ***
use ride-hailing
Perceived control/confidence to <--> Perceived ease of use easy of ride- 0.990 0.788 0.046 21.606 ***
use ride-hailing hailing
Perceived control/confidence to <--> Age 0.050 0.101 0.012 4.086 ***
use ride-hailing
Perceived safety risks of ride- <--> Gender 0.064 0.097 0.018 3.530 ***
hailing
Perceived safety risks of ride- <--> Age 0.057 0.099 0.015 3.871 ***
hailing
Perceived safety risks of ride- <--> Education 0.050 0.088 0.013 3.778 ***
hailing
Gender <--> Age -0.031 -0.142 0.006 -5.093 ***
Gender <--> Education -0.035 -0.162 0.006 -6.044 ***
Age <--> Income-group 0.022 0.109 0.005 4.486 ***
Age <--> Education 0.119 0.637 0.006 19.411 ***
Education <--> Car ownership 0.017 0.113 0.003 5.359 ***
Education <--> Income-group 0.048 0.244 0.005 9.726 ***
Income-group <--> Car ownership 0.067 0.404 0.005 13.074 ***
Notes: ß = Unstandardized regression estimates, b = standardized regression estimates; cov =covariance, corr =correlation;
*** P-value < 0.01. Correlation among error terms (e) depicted in the path diagram in Fig 3 are all statistically significant,
but not included in this table. Model fit indices: χ2 = 2601.315, df = 671, Normed-χ2 =3.876, p < 0.001; RMSEA = 0.049; NFI =
0.910; CFI = 0.921; IFI = 0.922; TLI = 0.904.
References
Acheampong, R. A. 2020. Spatial structure, intra-urban commuting patterns and mode choice:
Analyses of relationships in the Kumasi Metropolis, Ghana. Cities.
Acheampong, R. A., & Siiba, A., 2019. Modelling the determinants of car-sharing adoption intentions
among young adults: the role of attitude, perceived benefits, travel expectations and socio-
demographic factors. Transportation, 1-24.
Acheampong, R. A., & Cugurullo, F., 2019. Capturing the behavioural determinants behind the
adoption of autonomous vehicles: Conceptual frameworks and measurement models to predict
public transport, sharing and ownership trends of self-driving cars. Transportation research part F:
traffic psychology and behaviour, 62, 349-375.
38
Adarkwa, K. K., & Poku-Boansi, M. (2011). Rising vehicle ownership, roadway challenges and traffic
congestion in Kumasi. A.K. Kwafo (Ed.), Future of the tree: Towards growth and development of
Kumasi, 128-152. KNUST University Press, Kumasi Ghana
Agyemang, E., 2017. Mode choice for long distance trips: Evidence from the Greater Accra
Metropolitan Area of Ghana. Journal of Transport Geography, 64, 150-157.
Ajzen, I. 1991.The theory of planned behaviour. Organ. Behav. Hum. Decis. Process. 50, 179–211
Ackaah Jnr, E. K. D.,, W., Aprimah, B. A., & Adjei, E. 2020. Understanding demographics of ride-
sourcing and the factors that underlie its use among young people. Scientific African, e00288.
Alemi, F., Circella, G., Handy, S., and Mokhtarian, P., 2018a. What influences travelers to use Uber?
Exploring the factors affecting the adoption of on-demand ride services in California. Travel
Behaviour and Society, 13, 88-104.
Alemi, F., Circella, G., Mokhtarian, P., & Handy, S. 2018b. Exploring the latent constructs behind the
use of ridehailing in California. Journal of choice modelling, 29, 47-62.
American Public Transportation Association 2016. Shared mobility and the transformation of
publictransit. Report, American Public Transportation Association. Retrieved from
https://www.apta.com/resources/reportsandpublications/ Documents/APTA-Shared-Mobility.pdf,
Accessed date: 24 January 2018
Amoh-Gyimah, R., and Aidoo, E. N., 2011. Mode of transport to work by government employees in
the Kumasi metropolis, Ghana. Journal of Transport Geography, 31, 35-43.
Anderson, D. N. 2014. “Not just a taxi”? For-profit ridesharing, driver strategies, and
VMT. Transportation, 41(5), 1099-1117.
Bardhi, F., and Eckhardt, G. M., 2012. Access-based consumption: The case of car sharing. Journal of
Consumer Research, 39, pp. 881-898. http://dx.doi.org/10.1086/ 666376.
Barnes, S. J., and Mattsson, J., 2016. Understanding current and future issues in collaborative
consumption: A four-stage Delphi study. Technological Forecasting & Social Change, 104, pp. 200–
211.
Byrne, B. M. (2013). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic concepts,
applications, and programming. Psychology Press.
Birago, D., Mensah, S. O., and Sharma, S., 2017. Level of service delivery of public transport and
mode choice in Accra, Ghana. Transportation research part F: traffic psychology and behaviour, 46,
pp. 284-300.
Brown, A. E., 2018. Ridehail Revolution: Ridehail Travel and Equity in Los Angeles, PhD thesis URL:
https://escholarship.org/uc/item/4r22m57k accessed 15/10/2109
Burkhardt, J. and Millard-Ball, A., 2006. Who’s attracted to car-sharing? Transportation Research
Record: 1986(1), pp. 98–105.
Cheung, R., and Vogel, D., 2013. Predicting user acceptance of collaborative technologies: an
extension of the technology acceptance model for e-learning. Computers and Education, 63, pp.
160–175.
Circella, G., & Alemi, F., 2018. Transport Policy in the Era of Ridehailing and Other Disruptive
Transportation Technologies. In Advances in Transport Policy and Planning (Vol. 1, pp. 119-144).
Academic Press.
39
Clark, B., Chatterjee, K., and Melia, S., 2016. Changes to commute mode: The role of life events,
spatial context and environmental attitude. Transportation Research Part A: Policy and
Practice, 89, 89-105.
Clewlow, R. R. and Mishra, G. S., 2017. Disruptive Transportation: The Adoption, Utilization, and
Impacts of Ride-Hailing in the United States. Institute of Transportation Studies, University of
California, Davis, Research Report UCD-ITS-RR-17-07.
file:///C:/Users/r72323ra/Downloads/2017_UCD-ITS-RR-17-07.pdf
Conway, M., Salon, D. and King, D. A., 2018. Trends in Taxi Use and the Advent of Ride-hailing, 1995
– 2017 : Evidence from the US National Household Travel Survey. Urban Science, 2(79), pp. 1–23.
doi: 10.3390/urbansci2030079.
Degele, J., Gorr, A., Haas, K., Kormann, D., Krauss, S., Lipinski, P., Tenbih, M., Koppenhoefer, C.,
Fauser, J. and Hertweck, D., 2018, June. Identifying E-scooter sharing customer segments using
clustering. In 2018 IEEE International Conference on Engineering, Technology and Innovation
(ICE/ITMC) (pp. 1-8). IEEE.
Department of Urban Roads (2004) Urban Transport Planning and Management Studies for Kumasi
report. Volume I
Dias, F.F., Lavieri, P.S., Garikapati, V.M., Astroza, S., Pendyala, R.M. and Bhat, C.R., 2017. A
behavioral choice model of the use of car-sharing and ride-sourcing
services. Transportation, 44(6), pp.1307-1323.
Dias, F. F., Lavieri, P. S., Kim, T., Bhat, C. R., & Pendyala, R. M., 2019. Fusing multiple sources of data
to understand ride-hailing use. Transportation Research Record, 2673(6), 214-224.
Doppelt, L., 2018. Need A Ride? Uber Can Take You (Away From Public Transportation). Masters
Thesis, Georgetown University. URL:
https://repository.library.georgetown.edu/handle/10822/1050860 accessed 15/10/2019
Dong, Y., Wang, S., Li, L., & Zhang, Z. (2018). An empirical study on travel patterns of internet based
ride-sharing. Transportation research part C: emerging technologies, 86, 1-22.
Edelman, B. G., & Geradin, D., 2015. Efficiencies and regulatory shortcuts: How should we regulate
companies like Airbnb and Uber. Stan. Tech. L. Rev., 19, 293.
Efthymiou, D., Antoniou, C., & Waddell, P., 2013. Factors affecting the adoption of vehicle sharing
systems by young drivers. Transport policy, 29, 64-73.
Esson, J., Gough, K. V., Simon, D., Amankwaa, E. F., Ninot, O., & Yankson, P. W., 2016. Livelihoods in
motion: Linking transport, mobility and income-generating activities. Journal of Transport
Geography, 55, 182-188.
Ettema, D., 2018. Apps, activities and travel: an conceptual exploration based on activity
theory. Transportation, 45(2), 273-290.
Featherman, M.S., and Pavlou, P.A., 2003. Predicting e-services adoption: a perceived risk facets
perspective. Int. J. Hum. Comput. Stud. 59(4), pp. 451–474.
Flamm, B., 2009. The impacts of environmental knowledge and attitudes on vehicle ownership and
use. Transportation research part D: transport and environment, 14(4), 272-279.
Giddy, J. K., 2019. The influence of e-hailing apps on urban mobilities in South Africa, African
Geographical Review, 16, pp. 1–13. doi: 10.1080/19376812.2019.1589732.
40
Gomez-Morantes, J. E., Heeks, R., and Duncombe, R., 2019, May. A Multi-level Perspective on Digital
Platform Implementation and Impact: The Case of EasyTaxi in Colombia. In International
Conference on Social Implications of Computers in Developing Countries (pp. 195-206). Springer,
Cham.
Grahn, R., Harper, C. D., Hendrickson, C., Qian, Z., and Matthews, H. S. 2019. Socioeconomic and
usage characteristics of transportation network company (TNC) riders. Transportation, 1-21.
Graehler M, Mucci A, Erhardt G 2019. Understanding the Recent Transit Ridership Decline in Major
US Cities: Service Cuts or Emerging Modes? Transportation Research Board 98th Annual Meeting,
Washington, DC, January
Haddad, E.A., Vieira, R.S., Jacob, M.S., Guerrini, A.W., Germani, E., Barreto, F., Bucalem, M.L. and
Sayon, P.L., 2019. A socioeconomic analysis of ride-hailing emergence and expansion in São Paulo,
Brazil. Transportation Research Interdisciplinary Perspectives, p.100016.
Hall, J. D., Palsson, C. and Price, J., 2018. Is Uber a substitute or complement for public transit ?
Journal of Urban Economics, 108, pp. 36–50. doi: 10.1016/j.jue.2018.09.003.
Hampshire, R. C., Simek, C., Fabusuyi, T., Di, X., & Chen, X., 2018. Measuring the impact of an
unanticipated suspension of ride-sourcing in Austin, Texas (No. 18-03105).
Hart, J., 2016. Ghana on the go: African mobility in the age of motor transportation. Indiana
University Press.
Henao, A. and Marshall, W. E., 2019. The impact of ride hailing on parking (and vice versa). The
Journal of Transport and Land Use, 12(1), pp. 127–147. http://dx.doi.org/10.5198/jtlu.2019.1392.
Hess, S., Shires, J., & Jopson, A., 2012. Accommodating underlying pro-environmental attitudes in a
rail travel context: application of a latent variable latent class specification. In Working paper,
institute for transport studies. University of Leeds.
Hoffmann, K., Ipeirotis, P., & Sundararajan, A. 2016 . Ridesharing and the use of public
transportation.
https://pdfs.semanticscholar.org/1461/e4b58fc0a1ddce2d033bb0ebc99361df5de6.pdf
Hong, S. J., 2017. Assessing economic value of reducing perceived risk in the sharing economy: the
case of ride- sharing services. Journal of Public Transportation, 13(4), 61–72.
Hubert, M., Blut, M., Brock, C., Backhaus, C., & Eberhardt, T., 2017. Acceptance of smartphone‐
based mobile shopping: Mobile benefits, customer characteristics, perceived risks, and the impact
of application context. Psychology & Marketing, 34(2), 175-194.
Juschten, M., Ohnmacht, T., Thao, V. T., Gerike, R., and Hössinger, R., 2019. Carsharing in
Switzerland: identifying new markets by predicting membership based on data on supply and
demand. Transportation, 46(4), 1171-1194.
Kline, R. 2016. Principles and practice of structural equation modeling (Fourth edition.). New York:
The Guilford Press.
Kodransky, M., & Lewenstein, G. (2014). Connecting Low-Income People to Opportunity with Shared
Mobility. Report produced for Living Cities. New York, NY: Institute for Transportation &
Development Policy. URL: https://livingcities.s3.amazonaws.com/resource/284/download.pdf
Accessed 22/10/2019
Kufuor K. O., 2019. Uber in Ghana: markets and institutions in the emergence of ridesharing taxis.
URL
41
https://repository.uel.ac.uk/download/a0e08701204d88196de3ab3b3a881927e477c444347e027
37051f617cb78c827/505076/7493.pdf accessed 08/10/2109.
Lavieri, P. S., & Bhat, C. R. (2019). Investigating objective and subjective factors influencing the
adoption, frequency, and characteristics of ride-hailing trips. Transportation Research Part C:
Emerging Technologies, 105, 100-125.
Lewis, E. O. C., and MacKenzie, D. 2017. UberHOP in Seattle: Who, Why, and How?. Transportation
Research Record, 2650(1), 101-111.
Lindsay, G., 2017. What if Uber kills off public transport rather than cars? Accessed from
https://www.theguardian.com/sustainable-business/2017/jan/13/uber-lyft-cars-public-transport-
cities-commuting. [Retrieved May 18, 2017].
Mokhtarian, P. L., 2002. Telecommunications and travel: The case for complementarity. Journal of
industrial ecology, 6(2), 43-57.
Nie, Y. M. (2017). How can the taxi industry survive the tide of ridesourcing? Evidence from
Shenzhen, China. Transportation Research Part C: Emerging Technologies, 79, 242-256.
Okyere, D. K., Poku-Boansi, M., & Adarkwa, K. K., 2018. Connecting the dots: The nexus between
transport and telecommunication in Ghana. Telecommunications Policy, 42(10), 836-844.
Park, C.K., Kim, H.J., and Kim, Y.S., 2014. A study of factors enhancing smart grid consumer
engagement. Energy Policy 72, pp. 211–218.
Prieto, M., Baltas, G., & Stan, V., 2017. Car sharing adoption intention in urban areas: What are the
key sociodemographic drivers?. Transportation Research Part A: Policy and Practice, 101, 218-227.
Pham, A., Dacosta, I., Jacot-Guillarmod, B., Huguenin, K., Hajar, T., Tramèr, F., Gligor, V. and Hubaux,
J.P., 2017. Privateride: A privacy-enhanced ride-hailing service. Proceedings on Privacy Enhancing
Technologies, 2017(2), pp.38-56.
Rayle, L., Dai, D., Chan, N., Cervero, R., and Shaheen, S., 2016. Just a better taxi? A survey-based
comparison of taxis, transit, and ridesourcing services in San Francisco. Transport Policy, 45, 168-
178.
Rogers, B., 2015. The social costs of Uber. U. Chi. L. Rev. Dialogue, 82, 85.
Saadi, I., Wong, M., Farooq, B., Teller, J., & Cools, M., 2017. An investigation into machine learning
approaches for forecasting spatio-temporal demand in ride-hailing service. arXiv preprint
arXiv:1703.02433.
Sadowsky, N. and Nelson, E., 2017. The Impact of Ride-Hailing Services on Public Transportation Use:
A Discontinuity Regression Analysis. Economics Department Working Paper Series. 13.
https://digitalcommons.bowdoin.edu/econpapers/13
Schaller, B. (2018). The new automobility: Lyft, Uber and the future of American cities.
http://www.schallerconsult.com/rideservices/automobility.pdf.
Schwanen, T., 2015. Beyond instrument: smartphone app and sustainable mobility. European
Journal of Transport and Infrastructure Research, 15(4).
Shaheen, S. A., and Cohen, A. P., 2013. Carsharing and personal vehicle services: worldwide market
developments and emerging trends. International journal of sustainable transportation, 7(1), 5-34.
Simmons, R. O. B., 2018. Disruptive Digital Technology Services: The Case of Uber Car Ridesharing in
Ghana Url https://aisel.aisnet.org/amcis2018/OrgTrasfm/Presentations/11/ access 08/10/2109.
42
Tan, G.W.H., et al., 2014. Predicting the drivers of behavioral intention to use mobile learning: a
hybrid SEM-Neural Networks approach. Comput. Hum. Behav. 36, 198–213.
Venkatesh, V., & Davis, F. D. 2000. A theoretical extension of the technology acceptance model: Four
longitudinal field studies. Management science, 46(2), 186-204.
Wang, D., and Law, F. Y. T., 2007. Impacts of Information and Communication Technologies (ICT) on
time use and travel behavior: a structural equations analysis. Transportation, 34(4), 513-527.
Wang, Y., Gu, J., Wang, S., & Wang, J. 2019. Understanding consumers’ willingness to use ride-
sharing services: The roles of perceived value and perceived risk. Transportation Research Part C:
Emerging Technologies, 105, 504-519.
Ward, J. W., Michalek, J. J., Azevedo, I. L., Samaras, C., and Ferreira, P. 2019. Effects of on-demand
ridesourcing on vehicle ownership, fuel consumption, vehicle miles traveled, and emissions per
capita in US States. Transportation Research Part C: Emerging Technologies, 108, 289-301.
Yan, X., Levine, J., & Zhao, X. (2019). Integrating ridesourcing services with public transit: An
evaluation of traveler responses combining revealed and stated preference data. Transportation
Research Part C: Emerging Technologies, 105, 683-696.
43

Mobility-On-Demand: An Empirical Study of Internet-Based Ride-Hailing Adoption Factors, Travel Characteristics and Mode Substitution Effects

Uploaded by

Copyright:

Available Formats

Mobility-On-Demand: An Empirical Study of Internet-Based Ride-Hailing Adoption Factors, Travel Characteristics and Mode Substitution Effects

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mobility-On-Demand: An Empirical Study of Internet-Based Ride-Hailing Adoption Factors, Travel Characteristics and Mode Substitution Effects

Uploaded by

Copyright:

Available Formats

Authors accepted version of original article to appear in Transportation Research Part C: Emerging Technologies

Mobility-on-demand: An empirical study of internet-based ride-hailing

Key words: Mobility-on-demand; platform mobility; ride-hailing; travel behaviour; mode-

Today, internet-based on-demand mobility services commonly known as ride-hailing are

2 The rise of platform on-demand mobility services

3 Understanding ride-hailing adoption—A conceptual model and underlying

Fig 1: Conceptual model of on-demand ride-hailing mobility service adoption

Furthermore, travel choices have environmental consequences. Indeed, environmental

Another related concept that is considered relevant is perceived behavioural control.

4.2 Survey questionnaire design

Moreover, to understand mode substitution and complementarity effects, the survey

4.3 Data collection

4.4 Structural equation modelling

relatively larger proportion of youthful population. The age-groupings of the survey

Table 1: Summary of characteristics of the survey respondents.

Gender Female: Male 48%:52%

Age-groups 18-24 18%

Coupling Couple: Single: Divorced/Widowed 45%:53%:2%

Have children? Yes: No 42%:58%

Children in school (KG- Yes: No 84%:16%

Education Basic school (Primary + Junior High School) 16%

Employment Employed (Full-time) 50%

Income (n =812) ≤GH¢1,050 55%

Car Ownership/Use I have a car 15%

Driver’s Licensure Valid License: Learner, provisional: No 32%:5%:63%

5.2 Modelling ride-hailing adoption factors

5.2.1 Descriptive summary of responses to latent concepts

Openness to innovation, technology attitude and ICT use in travel

Car ownership/use attitudes, car-dependency and environmental values

I like trying things that are new and different 3 5 19 48 25

5.2.2 Overview of ride-hailing use and user characteristics

Finally, among the socio-demographic factors, educational attainments, income-group,

5.3 Characteristics of on-demand ride-hailing trips

5.3.1 Purpose of trips and reasons for ride-hailing

Fig 5: Travel costs associated with reference ride-hailing trips

5.3.2 Ride-hailing travel characteristics—journey start times, travel time, vehicle

from Weija to Ministries in Accra, covering approximately 20km.

5.3.3 Mode substitution effect of ride-hailing and multi-modality

6 Discussion and implications of findings

7 Conclusion, limitations and further work

use of ridehailing in California. Journal of choice modelling, 29, 47-62.

You might also like