Transportation Research Part A 132 (2020) 273-289

Transportation Research Part A 132 (2020) 273–289
Contents lists available at ScienceDirect
Transportation Research Part A

journal homepage: www.elsevier.com/locate/tra
Towards evening paratransit services to complement scheduled

T
public transport in Cape Town: A driver attitudinal survey of
alternative policy interventions
Christopher Plano , Roger Behrens, Mark Zuidgeest
⁎
Centre for Transport Studies, Department of Civil Engineering, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa
ARTICLE INFO ABSTRACT
Keywords: Cities across the Global South are seeking to improve public transport services, but the presence
Paratransit of large numbers of independent paratransit operators complicate reforms. City officials often
Cape Town seek to eliminate paratransit services in favor of scheduled services; but such reforms have
Public transport reform proven impossible to achieve. Cape Town is one such city, whose planning officials are now
Modal integration
reconsidering the approach to ‘industry transition’. Realizing that the transition and operating
Stakeholder engagement
Stated choice
costs of new scheduled services are unsustainable, the City has acknowledged a continued role for
paratransit services. The question this raises is how paratransit should be integrated and, more
specifically, how paratransit should complement scheduled services at transfer points. This paper
reports on findings from a stated choice survey administered to paratransit drivers to understand
opinions on potential reforms that the City might use to address issues of mismatched service
span and long off-peak headways in service provided by paratransit operators. The research seeks
to assist operators in sharing their perspectives outside of a City-structured engagement process
and to assist City officials in understanding what reform paths will be most feasible based on
driver acceptability and cost to the City. Results show that alternatives to the corporatization of
paratransit services are viable and appear to be more affordable.
1. Background
In many of the cities of the Global South, large numbers of independent operators provide much of the public transport (PT)
services. Local names for these services range from jeepney (Philippines) and trotro (Ghana) to car rapides (Senegal) and matatu
(Kenya), reflecting the wide range of contexts within which relatively similar services have emerged (Behrens et al., 2016; Cervero,
1985, 2000; Kumar & Diou, 2010; Lau, 1997; Saddier & Johnson, 2018). These services are sometimes referred to as ‘informal
transport’ (Cervero, 2000; Golub et al., 2009); however, in many cases this is a misnomer as operators are licensed or operate as
registered members of legally recognized associations or cooperatives (Behrens, McCormick, et al., 2016). Instead, though the term
was first used in the United States to describe services that are unscheduled and complement fixed PT (Kirby et al., 1974; Saltzman,
1973), paratransit is a better term and is defined by Schalekamp & Behrens (2010) as unscheduled PT provided using small vehicles
such as mini- or midibuses via unregistered, cash-based businesses.
In South Africa, paratransit is referred to as minibus-taxis (MBTs), which are mostly 15-seater vans. As in many countries,
operator decision-making is divided between two individuals, an owner and a driver. The owner provides the vehicle and is issued an
⁎
Corresponding author.
E-mail addresses: PLNCHR004@myuct.ac.za (C. Plano), roger.behrens@uct.ac.za (R. Behrens), mark.zuidgeest@uct.ac.za (M. Zuidgeest).
https://doi.org/10.1016/j.tra.2019.11.015
Received 26 March 2019; Received in revised form 4 October 2019; Accepted 15 November 2019
Available online 27 November 2019
0965-8564/ © 2019 Elsevier Ltd. All rights reserved.
C. Plano, et al. Transportation Research Part A 132 (2020) 273–289
Table 1
Summary of positive and negative elements of paratransit services.
Positive Negative
• Short access distance/high coverage (Hitge & Vanderschuren, 2015) • Poorly maintained vehicles (Cervero, 2000)
• Frequent (Cervero, 2000) • Poor off-peak frequency (McCormick et al., 2016)
• Greater security in smaller vehicle (Cervero, 2000) • Aggressive driving stemming from insecurity of pay (McCormick et al., 2016)
• Customer service, such as older passengers assisted with boarding/ • Overloading of passengers (Cervero, 2000)
alighting (own observations) • Passengers unable to board full vehicles near terminals (McCormick et al.,
• Demand-responsive (McCormick et al., 2016) 2016)
• Job creation for low-skilled young men (Schwanen, 2018) • Informal driver employment results in unpredictable income, job insecurity,
• Subsidy-free (McCormick et al., 2016) and no benefits (McCormick et al., 2016)
operating license that dictates the specific route on which the vehicle is permitted, as well as which terminals can be used for loading
(McCormick et al., 2016; TDA, 2014). Drivers determine the operational specifics of the service, deciding when to provide service.
Drivers collect fare revenue and either pay owners a rental fee, or ‘target’, each day or split the money with the vehicle owner on a
commission basis, where typically 70 percent accrues to owners and 30 percent to drivers (McCormick et al., 2016). A third in-
dividual called a ‘gaartjie’ or conductor is, if present, paid from the driver’s earnings and is responsible for collecting fares on board.
Owners must belong to an association and the association must approve any application for an operating license. Associations do
provide some regulatory control of the industry by coordinating service at ranks, though mainly serve the interests of owners rather
than drivers or passengers (Wilkinson, 2008).
The MBT industry provides critical links for residents of South African cities, but often this benefit is overlooked by other road
users who see only reckless driving. Many of the positive and negative elements of paratransit in South Africa are similar to those in
other countries, which are summarized in Table 1. In South Africa, most owners in the industry are Black African or Coloured,
historically disenfranchised groups who were often restricted from investing in other businesses; MBT businesses provide a generally
low barrier to market entry and therefore a path to economic empowerment (Barrett, 2003; Woolf & Joubert, 2013). In a survey of
MBT owners in Pietermaritzburg, South Africa, Magubane & Manicom (2003) found that 84 percent were Black African and 62
percent derived their sole incomes from the business.
1.1. South African MBT history
The characteristics of the MBT industry are rooted in more general market conditions that are common in other countries, but also
to the specific context of South Africa and its complicated past. An early version of the MBT industry developed by providing rides,
for a small fee, to passengers unable to board over-full state-owned PT that instituted segregated carriages that heavily favored
capacity for whites (Woolf & Joubert, 2013).
Transport challenges grew with the implementation of spatial segregation stemming from the Group Areas Act of 1950 (Mesthrie,
1993) which forced Black African and Coloured South Africans to the periphery of cities (Clark & Crous, 2002). While subsidized
transport services were provided by the state to allow workers to travel to places of employment, other transport needs were ignored
(Clark & Crous, 2002). The MBT industry grew to serve community needs for non-commute transport and provide an alternative to
subsidized bus and rail services that failed to provide adequate services (Bruun et al., 2016). As service became increasingly expensive to
provide, service coverage and frequencies declined (Barrett, 2003), opening the door to MBT industry growth. Exemplifying Gwilliam’s
(2008) regulatory cycle, competition from MBTs led to calls for regulation of the industry (McCormick et al., 2016).
In 1977, the Road Transportation Act enshrined into law the Van Breda Commission’s recommendations from the previous year
that effectively deregulated paratransit market entry by requiring that only vehicles carrying nine or more seats need acquire an
operating permit, when previous legislation defined this number as more than four seats (Bruun et al., 2016; Schalekamp & Behrens,
2010). Thus, so-called ‘combi-taxis’ were able to operate legally for the first time. However, it remained difficult to obtain a permit so
many larger vehicle operators continued to provide services illegally (Schalekamp & Behrens, 2010).
Competition regulation policy then underwent a pendulum swing: starting with the Welgemoed Commission which recommended
the protection of incumbent bus companies by the phasing out of ‘combi-taxis’; moving to a National Transport Policy Study which
recommended granting operating permits to illegal ‘minibus-taxis’ carrying 16 passengers; and ending with a Transport Deregulation
Act of 1988 which precipitated an exponential increase in the issuing of operating permits (Schalekamp & Behrens, 2010; Bruun
et al., 2016). The South African Black Taxi Association (SABTA), the voice of the MBT industry, warned prior to the Act’s passage that
deregulation would be problematic while also suggesting that deregulation was a government scheme to destroy the industry after
operators had begun to build wealth (Woolf & Joubert, 2013). Some warning signs were already present since the dissolution of the
railway police in 1986. The Road Transportation Act of 1977 had tasked the force with enforcing PT permits, and with their dis-
solution, little enforcement was occurring (Woolf & Joubert, 2013). As a result, taxi operators took matters into their own hands to
limit competition from illegal operators; often this self-regulation led to violence (Woolf & Joubert, 2013). Indeed, SABTA’s warning
proved prescient. Between 1984 and 1989, deregulation led to a 2500 percent increase in issued MBT permits (Khosa, 1992) and by
1989 MBTs were carrying the largest proportion of PT passengers (Schalekamp & Behrens, 2010). The violence that had been
occurring since 1986 increased dramatically as the industry reached saturation, eventually earning the name ‘taxi wars’ (Barrett,
2003).
274
With an obvious need for intervention, the National Taxi Task Team was formed by the post-apartheid government in 1994 to
consider solutions. Recommendations made in the White Paper on National Transport Policy of 1996 included formalizing MBTs
through the formation of associations and shifting to a route-based licensing structure that was intended to provide clearly defined
and separate operating areas for these associations (Schalekamp & Behrens, 2010; Woolf & Joubert, 2013). To promote formalization,
the government introduced the Taxi Recapitalization Program (Wilkinson et al., 2009). While it was originally intended to promote
formalization of business and employment practices, it effectively became a fleet renewal program that offered a partial capital
subsidy for new vehicles contingent on scrapping older vehicles as defined in the regulations (Wilkinson et al., 2009). While the taxi
recapitalization program has resulted in the removal of old and often unsafe vehicles, the impact of the program has been more
limited because the capital subsidy covers only a quarter of new vehicle cost (Schalekamp & McLachlan, 2016) and in general, limited
progress had been made in formalizing the MBT industry.
1.2. Cape Town context and reform
Because of limited progress in PT reform, a need for improvement remains. High unemployment and low wages result in a large
number of captive PT users (Clark & Crous, 2002) whose livelihoods depend on having efficient and reliable PT. Many of these
individuals live on the periphery because of apartheid spatial planning but also because more recent government housing policy has
favored the production of large numbers of housing units which necessitates construction on inexpensive land often located far from
desirable locations and places of employment (Bruun et al., 2016). Unfortunately, the PT system which might help address these
spatial problems is fragmented. Because multiple modes operate on independent fare payment systems, passengers avoid intermodal
transfers to save paying multiple fares (Bruun et al., 2016). As a result, operators duplicate each other’s routes. Further complication
stems from the multiple levels of government involved in PT provision, with national government operating rail, provincial gov-
ernment contracting for regional bus service, and city government contracting for MyCiTi Bus Rapid Transit (BRT) service (Clark &
Crous, 2002; TDA, 2018a). MBT regulation is under the purview of provincial government, but City of Cape Town officials are the de
facto authority for the approval or amendment of operating licenses (TDA, 2014).
Operationally, PT is difficult to provide in Cape Town because commuting distances are long, demand is highly peaked, and there
is intense competition between modes (Clark & Crous, 2002). Metrorail has traditionally served a large proportion of passenger
demand because it is well-integrated into the built environment (Clark & Crous, 2002), though in recent years ridership has declined
significantly with the increase in cancelled trains, delays, and other issues that impact heavily on reliability (De Klerk, 2017). Road-
based PT passenger demand is served primarily by MBTs that carry 60 percent of passengers as estimated by the City of Cape Town
transport demand model (TDA, 2018a), though some of the declining rail ridership has likely increased this proportion.
To address deficiencies in PT generally and related to the MBT industry more specifically, and stemming from the announcement that
South Africa would host the 2010 FIFA World Cup, in 2006 the Integrated Rapid Public Transport Network program was launched by
national government to promote the implementation of BRT by replacing MBTs (Schalekamp & Behrens, 2010). In Cape Town, the plan was
to replace MBTs with a BRT trunk and feeder system aligned with this national program (National Department of Transport, 2006, 2007).
Branded MyCiTi, the new system would incorporate existing MBT drivers in the areas where BRT would replace them to ensure
that employment opportunities remained constant (Schalekamp & Behrens, 2013; Schalekamp & McLachlan, 2016; Siyongwana &
Binza, 2012). To do this, MBT operators were induced to form vehicle operating companies that would negotiate 12-year contracts
with the City for MyCiTi operations (Schalekamp & McLachlan, 2016). However, negotiations for these contracts were continually
delayed, forcing the City to use interim contracts to ensure services were running for the World Cup (Schalekamp & Behrens, 2010).
There were many reasons for delays in negotiation (for more details, see Schalekamp & Behrens (2010) and Schalekamp & McLachlan
(2016)), not least of which is the historical distrust of government among members of the MBT industry. As discussed in Section 1.1,
SABTA saw deregulation as a government attempt to destroy the industry; similar sentiments were felt during this reform process
where industry felt that city officials were not making enough guarantees of employment opportunities, providing enough ownership
opportunities, or acknowledging the important role of the industry in economic empowerment (Schalekamp & Behrens, 2010;
Schalekamp & McLachlan, 2016; Woolf & Joubert, 2013). In addition, initial engagement with operators was through the South
African National Taxi Council (SANTACO), which is only theoretically representative of the taxi industry. Through extensive in-
terviews with operators in Cape Town, it is clear that operators preferred to be engaged directly as each is essentially an individual
business owner only loosely connected to others through registered associations (Schalekamp & McLachlan, 2016).
While some Latin American BRT systems are run without an operating subsidy (Sandoval & Hidalgo, 2002), this is rare and the
MyCiTi system is no exception (TCT, 2015; Von der Heyden, Laing, Hastings, Koch, & Reddy, 2015). Unfortunately much of the
explanation lies in apartheid spatial planning and its persistent effects on land use segregation and sprawling development (Behrens &
Salazar Ferro, 2016). To recover costs while keeping fares affordable, passenger demand must be high throughout the day and across
the network. Low density development, unidirectional travel, and a high peak-to-base ratio (Cooke et al., 2018) in Cape Town
essentially guarantee that PT, outside of key trunk corridors, will require an operational subsidy. MyCiTi feeder services exemplify
the issue, operating at a significant loss (TCT, 2015).
Because MyCiTi has failed to meet expectations for cost recovery and because operator engagement has proven both difficult and
expensive, the City of Cape Town has acknowledged, through adoption of its Integrated Public Transport Network Business Plan, that
“the Phase 1 methodology is not viable or sustainable” and that MBTs will provide unscheduled feeder services to scheduled trunk
services within a hybrid (including both scheduled and unscheduled services) PT network (TDA, 2017). This aligns with earlier calls
for MBTs to act as complementary partners for PT provision (Behrens et al., 2016; Salazar Ferro et al., 2013). Within a trunk and
feeder arrangement, South American cities have experimented with a number of mechanisms to ensure a smooth transfer across
275
paratransit and scheduled modes. These generally fall into three categories: reward schemes, area licensing, and concessions (Behrens
et al., 2016). While these cases provide some insight into potential approaches for integrating paratransit as feeders into a multimodal
network, past attempts at wholesale adoption of technology from South America have proven problematic, suggesting it is critical to
use these cases only as guides (Salazar Ferro, Muñoz, & Behrens, 2015).
Cities in South Africa are also experimenting with MBT integration approaches. The City of Cape Town is working with existing taxi
associations in one area to form a Taxi Operating Company (TOC) that would be contracted to provide feeder services. While not
focused on feeder services per se, the eThekwini Municipality (Durban) is in the planning stages for the Public Transport Accelerated
Service Improvement Program (or Moja Cruise) that would pay drivers monthly for improved vehicle maintenance, driving behavior,
and customer service (Schalekamp & Klopp, 2018). Formalization is not an immediate goal of Moja Cruise, which sets it apart from
many reform approaches.
The goal of any reform effort is to obtain the best transport network possible for the least cost, both in monetary and social terms
(such as displaced MBT operators, reduced access to employment, etc.). Learning from the Phase 1 approach in Cape Town, it is clear
that government must meaningfully and extensively engage with MBT operators if changes are to be made to promote feeder service
provision (Schalekamp & Behrens, 2013), especially because any interventions in the industry, such as a change in fare payment system
from cash to cashless impacts the entire business model and relations between operators, labour and passengers (Schalekamp et al.,
2017). However, the operator perspective on specific interventions to achieve MBT feeder service remain unknown because no research
has been conducted around this question. This paper will describe the first half of a research project that attempts to provide the
paratransit operator perspective on interventions that are likely to encourage MBT services to be provided later and, in some cases, with
reasonably short headways. This paper describes the driver’s perspective while a future paper will describe the owner’s perspective.
2. Method
2.1. Study area
MBT operators are currently providing feeder services to trunk routes at the Mitchells Plain Public Transport Interchange (PTI), providing
an opportunity to test the feasibility of interventions to improve transfer quality in a future network by using a piece of the existing network
(Fig. 1). Located 27 km south-east of the Cape Town Central Business District (CBD), the PTI is one of the busiest in the City, serving more
than 20,000 passengers per day (TDA, 2018a). Transport services include local and line-haul PT provided by Metrorail, Golden Arrow Bus
Services, MyCiTi, and MBTs. The area around the PTI is also a destination in its own right for retail and employment at the Town Centre.
Another major retail and employment hub, the Liberty Promenade Mall, is located nearby and is linked by MBT service to the PTI.
Mitchells Plain was built in 1973 and is the largest Coloured township in Cape Town with nine neighborhoods of mostly low and middle
income residents (Smith & Hanson, 2003). According to the 2011 census, the total population of the area is approximately 310,000 of which
nearly 95 percent live in formal dwellings (Statistics South Africa, 2018). Because feeder-trunk-distributor schemes are often implemented
with feeders/distributors in low to middle income residential areas, Mitchells Plain is an appropriate location to use as a case study.
Previous research has assessed the level of complementarity between MBT feeder/distributor service and scheduled Metrorail,
MyCiTi, and Golden Arrow trunk service at the Mitchells Plain PTI (Behrens et al., 2017). The study assessed complementarity with
respect to MBT service span and frequency. These are common quality of service factors globally for scheduled PT services
(Transportation Research Board, 2013); for MBT services that operate without schedules, service frequency can be assessed using
vehicle departures from the main ranks (Behrens et al., 2017; Saddier, Patterson, Johnson, & Wiseman, 2017). In this case, issues
were found particularly related to service span, with MBT service ending too early to accommodate later trunk service arrivals, as
well as some indication of long wait times for MBT departures during off-peak times. While passenger satisfaction is reasonably high
for these feeder routes in general (Behrens et al., 2018), the City of Cape Town seeks to provide 18-hour service across the PT network
(TDA, 2014), which aligns with international standards for service quality (Transportation Research Board, 2013). In addition,
passengers surveyed regarding satisfaction of MBT service were sampled between the hours of 9 am and 1 pm, suggesting that these
passengers are less likely to be concerned by the lack of evening service (Behrens et al., 2018).
2.2. Survey method
In many cities, PT services are operated either directly by government authorities or by private operators under contract.
Adjusting frequencies, hours, and other service aspects is simply a matter of reallocating resources within the organization or re-
negotiating terms with contractors. However, in many cities of the Global South government cannot impose service level control over
paratransit operators who operate independent, commercially viable services. Therefore, to improve services, operators must be
willing to provide it. To measure this willingness to provide improved service under various scenarios, or interventions, and in the
absence of revealed data on such willingness, we used a stated choice (SC) survey. This is a method commonly used to determine
demand for various modes of transport, especially when governments are considering a major investment such as a new rail line
(McFadden, 1974). A SC survey provides data that, through choice modelling (Hensher et al., 2015), can indicate what proportion of
the traveling public will use a particular mode, allowing planners to complete a benefit-cost analysis of the investment.
For our purposes, we inverted the method. Rather than ask commuters whether they would drive, ride the bus, or take the train,
we asked MBT drivers how willing they would be to provide service. Drivers were the focus because they are the key operational
decision-makers (McCormick et al., 2016). The output is a probabilistic indication of driver willingness to provide service under
varying conditions, allowing government to understand which conditions are likely to convince drivers to provide the desired service.
276
Fig. 1. Location of Mitchells Plain PTI relative to Cape Town CBD. Source: Authors using City of Cape Town Open Data Portal, https://web1.
capetown.gov.za/web1/opendataportal/Default.
As discussed in an earlier paper (Plano et al., 2018), examples of measuring willingness to provide a service using SC methods is
limited, though a number of cases can be found in the conservation farming literature (Broch et al., 2013; Espinosa-Goded et al.,
2010; Schulz et al., 2014). Other examples include surveys to understand the preferences of healthcare service providers for varying
employment conditions (Scott, 2001; Scott et al., 2007; Ubach et al., 2003). In reviewing the academic literature, no examples could
be found where a SC survey was used to assess operator willingness to provide PT services, suggesting this research is a unique
addition to the literature. For lessons learned in designing and using this method for a novel application, see Plano et al. (2018).
Because the main issue identified by previous research was a mismatch in PT service spans, the SC survey asked MBT drivers
whether they would be willing to provide service for an additional three hours from 7 pm to 10 pm. The response format was ordered
(Definitely no, Probably no, Unsure, Probably yes, Definitely yes), which allowed estimation of both ordered and binary models. A
focus group was held with owners, drivers, and owner-drivers to determine factors important in a driver’s decision to provide service
(profit, security, and passenger loyalty) which informed the attributes included in the survey, listed in Table 2. The Farepay and
Headway attributes are not considered by drivers in status quo choice-making, but were included to understand willingness to
provide service under various interventions (Table 6) aimed at addressing the service complementarity defects described above.
Dummy attribute levels stemmed from the focus group (desire for a police officer instead of current private security staff) and design
of interventions, some of which include CFC (rather than cash fares currently) and/or headway conditions (rather than vehicles
leaving when full currently). Headway attribute levels were based on previous departure time data (Behrens et al., 2017) and service
planning standards literature. Attribute levels for the profit variable were determined through regular informal discussions with
Table 2
Attributes included in the SC survey.
Attribute Description Type Levels
Profit Total driver profit 7–10 pm (ZAR) Variable 30, 40, 50, 60, 70, 80
Security Security provided at the rank Dummy Security guards, Armed South African Police Services officer
Farepay Fare payment method used Dummy Cash, Cashless
Headway Requirement for vehicles to leave at specified Dummy None (vehicles leave when full), Vehicles leave every 20 min, Vehicles leave
intervals every 10 min
277
Fig. 2. Example choice card presented to respondents.
drivers on income and costs over five months, data exchange with others who have undertaken research on the MBT industry in South
Africa, and by calculation.
A small pilot was conducted with four individuals to test for understanding and to ensure attributes and levels were appropriate.
Pilot respondents were interviewed directly after the survey to check for understanding and to determine which attributes were
considered; as a result, the passenger loyalty attribute was removed. Because pilot respondents consistently chose no (yes) responses
when profit levels were low (high), profit attribute levels were narrowed in range to encourage greater trading.
Terminology used needed to be carefully designed to ensure respondents understood concepts as intended by researchers (Mangham,
Hanson, & McPake, 2009; Venter & Venkatesh, 2010). The focus group, informal discussions with drivers, and the pilot all served as an
opportunity to note how particular elements of the MBT industry were discussed. The informal data-gathering process, an appropriate
alternative to formal interviews (Gillham, 2000), fed directly into questionnaire design and terminology was later confirmed in a meeting
with the taxi association leadership. As these individuals were intimately involved with the particular context of the association from
which driver respondents would be sourced, but would not be taking the survey, they were instrumental in refining terminology.
In the full survey, respondents were presented with 12 choice cards and one scenario per choice card to maximize data collected
from the limited pool of respondents and to reduce respondent burden (Arentze et al., 2003; Caussade et al., 2005). Table 2 and Fig. 2
provide details on the final survey. Ngene was used to design the choice cards with a prior value of 0.01 for Profit and zero for all
others because pilot estimates were significant only for Profit. The final d-efficient design had a d-error of 0.39 compared to a
benchmark model of priors, whereby a lower error implies a more efficient design. At the conclusion of the SC portion of the survey,
respondents were asked to provide additional data such as age, daily profit, and attitudes towards various aspects of the business (e.g.
would you rather be paid a salary?). Surveys were administered by trained interviewers fluent in Afrikaans and/or isiXhosa and
conducted as much as possible inside drivers’ vehicles to provide some privacy within the rank environment.
Survey respondents were sourced from two taxi associations that provide feeder service to the Mitchells Plain PTI: Hazeldene
Shuttle Services (HSS); and Seventh Avenue Taxi Association (7ATA). HSS is led by two individuals and operates a consolidated route
with a flat fare of ZAR 7 (USD 0.54) while 7ATA is led by a committee, runs three routes, and has a slightly higher fare of ZAR 8 (USD
0.62)1. These associations were selected from 102 associations registered in Cape Town. Because this research looks specifically at
complementarity between scheduled trunk and unscheduled feeder services, only associations operating feeder services were eligible
to be included in the SC survey. Of all surveyed routes in Cape Town, 50 percent are estimated to be feeder routes (du Preez,
Zuidgeest, & Behrens, 2019). However, associations often operate more than one route; assuming each operates an average of two
routes, there are approximately 25 associations operating feeder services in the city. Despite this seemingly large sampling frame, the
1
Exchange rate from September 2017: ZAR 13 to 1 USD.
278
Table 3
Characteristics of respondents.
Characteristic HSS 7ATA
Total respondents 35 44
Percent male 92% 98%
Average age 53 years 44 years
Percent drivers (v. owner-drivers) 78% 100%
major limitation for research with the MBT industry is the contentious nature of MBT regulation and reform and the potential for
conflict, as discussed in Section 1. Researchers had an established relationship with the two associations surveyed from previous
research (mentioned and cited above) and a capacity-building program facilitated by a colleague. This provided a foundation of trust
that enabled this research. Previous research findings also provided understanding of complementarity issues that allowed pro-
gression to this solution-focused research. Because of the pressing issue surrounding the role of MBTs as feeders in a hybrid network
as a result of the city’s change in planning paradigm, researchers felt it was important to investigate potential solutions using the
positive working relationship with these particular associations rather than attempt to sample from all feeder associations. Therefore,
this paper makes no claim that results are representative of feeder associations in general and rather seeks to capitalize on a well-
understood case to inform future PT policy and planning.
Across the two associations, the driver and owner-driver population was estimated by association leadership to be 123 in-
dividuals. According to Johnson and Orme (2003), the survey design used here with 12 choice tasks, two alternatives, and six levels
for the Profit attribute and interested only in main effects requires 125 respondents. Therefore, the survey was intended as a census of
HSS and 7ATA drivers and owner-drivers. Unfortunately, time and budget constraints as well as the fluid nature of the MBT industry
made it practically impossible to reach all 123 individuals; after many days of survey administration, no additional respondents could
be found at any of the ranks.
There were a total of 79 respondents across both associations (Table 3); because one individual from 7ATA left all choice cards blank,
choice modelling included only 78 respondents while analysis of non-stated choice data included all 79 respondents. In total, approxi-
mately 78 (56) percent of HSS (7ATA) individuals were surveyed. Administration with HSS occurred in the third week of October 2017
and with 7ATA in the last week of November 2017. For more details on survey design and administration, see Plano et al. (2018).
3. Results
3.1. Choice modelling
Both binary and ordered logit models were estimated using an error component specification to account for the correlation in
unobserved elements across responses given by a particular individual (Abdel-Aty, Kitamura, & Jovanis, 1997; Revelt & Train, 1998).
The ordered logit model was estimated using aggregate data that included 889 total observations across 78 respondents (59 choice
tasks were left blank and were excluded). The binary logit model was estimated using 864 total observations across 78 respondents
with the same 59 choice tasks excluded, with an additional 25 Unsure responses also excluded. The respondent population for each
association individually is small; therefore, the aggregate data set was used to provide greater statistical power.
Estimation was completed using simulated maximum likelihood in the Biogeme software (Bierlaire, 2018). A binary logit model was
estimated for all five parameters (the attribute pertaining to time between departures at the rank, or headways, was coded to allow for
non-linear relationships between the three levels). An ordered logit model was estimated for all five parameters as well as the threshold
values τ that indicate the division between response categories on a latent utility scale. A series of models with various combinations of
parameters were estimated to explore outcomes; binary models produced significant estimates only for Profit, ASC, and sigma while the
ordered models resulted in a significant security attribute as well. Farepay and Headway parameter estimates were consistently in-
significant, which aligns with how respondents self-reported which attributes they considered, with only 27 (32) percent considering the
Farepay (Headway) attributes compared to 75 (67) percent for Profit (Security) attributes. This suggests attribute nonattendance.
After comparing model results, an ordered logit model with only the Profit and Security parameters was selected for further
analysis because the survey was designed as ordered and it included the attributes most often considered by respondents. Model
results are shown in Table 4. τ values calculated from delta estimates (τ2 = τ1 + δ2 and so on) provided by Biogeme are 0 (fixed),
0.936, 1.232, and 2.982. The best binary model included only the Profit attribute; for comparison, these results are shown in Table 5.
The lack of significance of the Farepay and Headway attributes can likely be attributed to the fact that they are non-standard
attributes; typically choice modelling attributes are chosen because respondents consider those factors when making the choice being
studied, and in this case the Profit and Security attributes adhere to this rule. However, the goal of the study was to assess policy
scenarios and required the addition of attributes that would reflect future scenarios and therefore some attribute levels for the
Farepay and Headway attributes were unusual for drivers to consider when making a choice to provide services. As a result, it is
reasonable that many drivers simply ignored them as evidenced by the self-reported data mentioned above.
Additionally, the significance of the sigma parameter indicates there is considerable influence of unobserved factors, but these
factors were not identified through the initial focus group used to design the survey. Some of these factors may be related to
individual characteristics, but as discussed below, for the policy goals of this research, knowledge of how individual characteristics
impact choice-making is not useful.
279
Table 4
Ordered model estimation results.
Statistic Value Parameter Estimate Robust s.e. Robust t-test Robust p-
value
Null LL -1988.793 ASC -1.58 0.621 -2.54 0.011
Final LL -846.67 Profit 0.0533 0.0075 7.11 0.000
AIC 1707.34 Security 0.406 0.175 2.32 0.020
BIC 1723.837 Sigma 3.69 0.483 7.63 0.000
Adj. Rho square 0.571 Delta2 0.936 0.236 3.96 0.000
Observations 889 Delta3 0.296 0.0833 3.56 0.000
Individuals 78 Delta4 1.75 0.254 6.89 0.000
Table 5
Binary model estimation results.
Statistic Value Parameter Estimate Robust s.e. Robust t-test Robust p-
value
Null LL -598.8792 ASC -3.36 1.50 -2.24 0.025
Final LL -295.9609 Profit 0.0831 0.0145 5.71 0.000
AIC 597.9218 Sigma 4.79 0.82 5.84 0.000
BIC 604.9919
Adj. Rho square 0.501
Observations 864
Individuals 78
Table 6
Description of interventions aimed towards achieving evening service provision.
Intervention Description Service quality defects addressed
Higher evening fare Association/drivers charge higher fares in the evening to allow drivers to operate profitably Service span
with lower demand. No cost to city.
Rank security SAPS officer stationed at the rank in the evening. Extra security could make drivers feel safe Service span
enough to operate later.
Transfer bonus Small amount paid per passenger that transfers from trunk to MBT feeder in off-peak, counted Service span
by intermodal cashless fare collection (CFC) technology.
Headway bonus Payment per vehicle trip that departs in less than specified headway monitored via vehicle Service span and headway
tracking devices.
Operating deficit payment Payment to cover operating deficit plus a small profit via MBT-only CFC technology. Service span, possibly small
headway effect
Off-peak contract Newly formed MBT operating company is contracted to provide off-peak service at specified Service span and headway
service quality levels monitored by CFC and vehicle tracking devices.
E-hailing Drivers use e-hailing app to serve passenger demand during off-peak periods via shared rides. Service span only, but may
reduce wait time
To achieve the goal of extending MBT service until 10 pm, the ordered logit model was used to calculate the probability of a driver
agreeing to provide evening off-peak service under two conditions; one with only the current contracted private security personnel at
the rank and one with an armed South African Police Service (SAPS) officer. The probability of a driver agreeing to provide service is
higher for a given profit level when a SAPS officer is at the rank. Exchange SAPS for private security, and the probability of drivers
agreeing to provide service is lower for the same amount of profit.
To obtain a 70% probability of drivers agreeing to provide service (combining the probabilities for being in Probably yes or Definitely
yes categories), MBT drivers and owner-drivers would require a SAPS officer at the rank and ZAR 61.04 (USD 4.69)2 profit from 7 pm to
10 pm. Without a SAPS officer, ZAR 68.66 (USD 5.28) profit would be required to maintain the same probability of a driver agreeing to
provide service. In both cases, there is a 29% chance of a driver responding Definitely yes and a 41% chance of responding Probably yes.
This difference in profit required to maintain driver willingness to provide service at the same level indicates the willingness to
accept (WTA) value of ZAR 7.62 (USD 0.59); in this case, it is the value placed on additional security at the rank. With only private
security, drivers feel less secure in the evening hours and demand ZAR 7.62 (USD 0.59) higher compensation over the three-hour
period to agree to provide service.
While driver pay is not based on hours worked, it is useful to compare in terms of hourly rates. Dividing the above profit values by
three indicates that with (without) SAPS, ZAR 20.35 (ZAR 22.89) per hour is required to encourage service provision while current
2
Exchange rate from September 2017: ZAR 13 to 1 USD.
280
50% n
structure Status Gender

Male 841
Female 48
Owner-driver 96
Driver 793
Commission (7ATA) 469

Payment
Target (HSS) 324
Owner-driver 96
26-35 136
36-45 109
Age
46-55 380
56-65 216
66-75 48
1-3 178
Household
4-6 602
size
7-9 97
10-12 12
0 662
1 95
incidents
Robbery
2 36
3 48
4 or more 36
0-1 561
2-3 205
Nights in
activities
4-5 63
6-7 60
All respondents 889

|
Definitely no Probably no Unsure Probably yes Definitely yes
Fig. 3. Impact of respondent characteristics on choice task response. Note: percentage calculations and n use observations rather than individuals.
minimum wage for all workers as specified by the National Minimum Wage Act of 2018 is ZAR 20 (National Minimum Wage Act, 2018).
This somewhat higher amount suggested by the survey could be due to a number of factors, including the security risk drivers perceive
both at the rank and along the route. SAPS may mitigate this risk at the rank but do little to reassure drivers on the route. Working
evening hours also requires foregoing family time or other activities that may lead drivers to demand additional compensation.
The survey asked additional questions such as respondent age and household size to allow description of the respondent popu-
lation, but also to determine if such characteristics impacted the choice results (Fig. 3). While additional choice modelling can be
done to determine parameters for sociodemographic characteristics, this information is of no practical use to the public authority;
surveying a MBT association’s drivers and paying varying amounts of incentives increases the industry transition workload for the
authority and, more importantly, is likely to cause contention among drivers. Therefore, a more basic analysis of impacts on choice
responses was done for interest rather than practical policy application and justifies use of the error component specification rather
than providing for random parameters that would allow additional information about heterogeneity.
Across all respondents, 57 percent of choice tasks elicited Definitely yes (38 percent) or Probably yes (19 percent) responses. The
unsure response was used only three percent of the time. A total of eight percent were Probably no and 32 percent were Definitely no.
Only four females were surveyed—three from HSS and one from 7ATA—and they are more likely to agree to provide evening
service than their male counterparts. Owner-drivers (all eight of whom are members of HSS) are less likely to drive later possibly
because all revenue collected is their own; compared to drivers, they have fewer expenses without a daily target or commission.
Within the pay structure comparison, owner-drivers are again the group least likely to provide additional evening service. Other
results related to pay structure seem counterintuitive, where drivers on a commission system (total earnings are used to pay fuel and
conductor and remaining amount is split 30 percent to the driver and 70 percent to the owner) might be expected to be less likely to
drive because of the unfavorable split, whereas drivers on the target system might be able to reach their target during the day and
keep all evening profit.
281
Owner-driver/driver age does not demonstrate a major impact on choice responses; those aged 36–45 years or over 65 are less
likely to drive. While older drivers may not want to drive longer hours perhaps because of fatigue, the results for younger drivers are
unexpected. A similar effect can be seen in household size, with only the 7–9 group showing major differences (much less likely to
drive) though this is a relatively small group. It might be expected that larger households would lead to greater pressure for earnings
from owner-drivers/drivers, though this was not the case.
The vast majority of respondents have not had a single robbery incident in the past year, which is surprising because focus group
participants almost unanimously agreed that safety and security were major issues in the area, particularly at night. Being cash-based
businesses, there is concern that drivers become targets at the end of the day if cash is not deposited. However, it may be that
perception of risk is greater than actual risk, or that robbery is not as problematic compared to other risks such as gang-related
violence. For the most part, choice task responses align with expectations that more robberies would reduce driver willingness to
drive into the evening. Observations in the 4 or more category counter this logic.
Individuals involved in activities more nights per week are less likely to agree to drive into the evening though it may be hard to
draw conclusions when such a large proportion of responses fall into the first two categories.
Overall, the data suggest that the proposition of later MBT service in the Mitchells Plain area is somewhat likely. The question we
turn to now is how drivers feel about each of the interventions that operationalize the security or monetary incentives.
3.2. Driver acceptability of interventions
The results of the survey and choice modelling were used to determine the acceptability, from the driver perspective, of various
interventions that would encourage them to drive later and therefore make service available to passengers until 10 pm. These
interventions are described in Table 6.
As much as possible, these interventions were informed by the literature. However, paratransit integration has only recently been
given the attention required to adequately inform PT reform efforts and the literature remains limited. Many of these interventions
may never have been attempted, though this is difficult to say unequivocally as much work is not documented. Regardless, inter-
ventions are typically limited to fleet renewal or professionalization of paratransit, limiting information for this research but high-
lighting the need for it (Salazar Ferro, 2015). Reliance on informally shared knowledge from experts in the field was an important
way to ensure that this set of interventions covered most, if not all, realistic options. To that end a number of South African para-
transit experts were consulted; their names are listed in the Acknowledgements.
All but one intervention aims to increase driver earnings to encourage service provision. Associations or drivers charging higher
evening fares would directly accomplish this with no involvement from the public authority. This has been suggested by Salazar Ferro
(2015) and has been independently implemented by associations in Cape Town on some routes, as in the case of Main Road Route
Taxi Association where posted fares indicate an increase for evening travel. A transfer bonus was used in Quito to reward paratransit
drivers bringing passengers to the BRT without requiring formalization or schedules (Behrens et al., 2016; Salazar Ferro et al., 2015).
To accomplish this, a CFC system is needed to count passengers qualifying for a transfer bonus, which in turn makes driver income
visible and taxable. A tracking device simply ensures drivers do not cheat and actually provide trips.
Rewarding drivers for providing specified headways directly addresses headway regularity issues as well as the service span issue.
Again, a tracking device ensures driver honesty and drivers would need to pay tax on at least the bonus money paid by the public
authority as these are required to be paid electronically or by cheque and are therefore traceable (Municipal Finance Management
Act, 2004). The Moja Cruise program intends to incentivize drivers to provide fixed headways in a similar way (Schalekamp & Klopp,
2018) and a concession system in Bogota used a similar concept of driver monitoring linked to a system of penalties and rewards
(Hernández & Mehndiratta, 2015). An operating deficit payment may also address the headway issue to some degree by ensuring
drivers a small profit regardless of passenger count, allowing them to make equivalent revenue with fewer passengers and therefore
shorter waiting times. Cashless fare collection will assess revenue while vehicle tracking will be necessary to determine fuel costs as
well as ensure driver honesty. Though instituted in a concession-like context, the Dakar bus renewal scheme incorporated a deficit
payment to compensate operators in the case of government denying a fare increase request (Kumar & Diou, 2010).
Under the off-peak contract reform, feeder services are provided by a fully formalized MBT sector likely with gross cost, per km
contracts as used in Phase 1 of MyCiTi as well as in George (Schalekamp & Klopp, 2018) and in other cities such as Bogota (Salazar Ferro
& Behrens, 2013). Drivers would be paid taxable salaries (SARS, 2017; Schalekamp & Klopp, 2018) and vehicles would be owned and
scheduled by the company to provide scheduled service. Cape Town intends to develop an integrated CFC system for the PT network,
suggesting that any service would require CFC (TDA, 2018b). Contract adherence would be monitored via vehicle tracking devices.
The e-hailing reform would not require public authority funding related to MBT operations beyond the cost of developing and
maintaining the app to be used by drivers. The expectation is that passenger demand would either increase or become more visible to
drivers and would increase their earnings enough to encourage them to drive later (Aftarobot, 2018). Similar to the operations of UberPool
for example, vehicles would be routed based on passenger boarding and alighting requests through the app. Similar to Uber in South
Africa, payment could be cash or through the app with either situation triggering payment of tax as amounts are visible through the app.
Because the Off-peak contract cost does not include costs for automatic fare collection (a separate contract following best
practice), the cost of CFC is excluded from other interventions to ensure comparability (Transfer bonus, Operating deficit payment).
Likewise, the cost of vehicle tracking is included in Transfer bonus, Operating deficit payment, and Headway bonus to compare to E-
hailing and Off-peak contract which include these costs inherently.
The one intervention that does not involve increasing driver income is where the public authority provides additional security via
a police officer. This intervention would simply increase driver perceptions of security which, through the initial focus group held to
282
assist in survey design, seemed to be a promising way to obtain evening service with no involvement in MBT operations.
Because some choice model parameter estimates were not significant, an alternative method, using the non-choice task survey
questions, was used to understand respondent preferences for interventions and their subcomponents. Each intervention has its own
set of necessary elements as described above; questions in the survey such as, “Is it fair for you to pay tax on your income?” are
therefore indications of how acceptable drivers find particular pieces of an intervention. For an overall intervention to be considered
acceptable to a driver, that driver must find each subcomponent acceptable, responding affirmatively to all questions. Table 7 shows
the survey questions that were linked to each intervention.
The two questions most often linked to interventions are willingness to pay tax and willingness to use a tracking device. Only the
Higher evening fare intervention does not involve these conditions because it involves minimal change to the operating environment.
Paying income tax relates in some interventions to incentive payments only, while in others includes fare revenue as well. E-hailing
and Off peak contract interventions involve a number of unique conditions because each requires major shifts from the current
operating environment. Rank security acceptability was determined from driver responses to an open-ended question requesting
them to identify what is most important to passengers; a large proportion responded with some variation on safety and these
proportions are indicative of drivers willing to accept a police officer at the rank.
The results of this analysis are shown in Table 8. Interventions are sorted from most to least acceptable by the first column.
Percentages listed in the Combined column indicate the percent of owner-drivers/drivers across both MBT associations that find that
intervention acceptable. Many of the subcomponents of interventions are necessary (such as a tracking device to prevent cheating)
and therefore these cannot be policy instruments to improve acceptability. However, the payment of tax can be adjusted through an
exemption. By removing the requirement that MBT drivers pay tax on earnings, the public authority can increase the acceptability of
some interventions dramatically as shown in the difference between Tax owed and Exempt columns.
Rank security stands out as most acceptable to drivers, which aligns with findings from the focus group used to design the survey;
the vast majority of participants supported having a police officer at the rank. This intervention requires funding for an officer’s time.
Two other interventions make it over the fifty percent mark overall: Higher evening fare and Headway bonus. The former requires no
public authority intervention while Headway bonus requires funding of incentive payments and the installation of tracking devices on
vehicles. After Rank security, there appear to be tiers of acceptability with Higher evening fare and Headway bonus together; then
Transfer bonus and Operating deficit payment (these figures are identical because each has the same subcomponents); followed by
Off-peak contract and E-hailing at the lower end of acceptability. Interventions are generally less acceptable among 7ATA re-
spondents than among HSS respondents.
The associations differ in acceptance of most interventions, but the largest difference occurs with E-hailing, where a spread of 35
percentage points separates the two associations. The smallest difference is with Off-peak contract. The difference between ac-
ceptability of Headway bonus and Transfer bonus/Operating deficit payment can be contributed entirely to the addition in the latter
of a CFC system. It is clear that there is more resistance to paying tax among HSS respondents than for those at 7ATA.
3.3. Intervention costs
Because one of the main concerns of the public authority is cost, the cost of each intervention to the public authority was
estimated. Operational cost data such as fuel usage was collected through informal discussions with drivers while self-reported
revenue data came from the survey. This information, combined with estimated evening passenger demand, was used to create a
simple MBT business model with parameters that could be manipulated for each intervention. With the estimated values for the profit
attribute in the choice model highly significant, the value of ZAR 68.66 (USD 5.28) was used as the threshold level of compensation
needed for service provision (to obtain a 70 percent chance of an owner-driver/driver agreeing to provide service); any shortfall
between this amount and that indicated by the business model is the potential cost to the public authority. These costs are intended to
be order of magnitude estimates rather than exact costs to provide an indication of what reforms may be most feasible overall.
Because the joint cost model aggregated HSS and 7ATA revenue and cost data, results for cost estimations are presented for both
associations. In other words, costs for each intervention in Table 9 and Fig. 4 include the costs to the public authority (or to the
passenger in the case of Higher evening fare) of the intervention to be rolled out to both associations. For Rank security for example,
this requires two SAPS officers because the associations operate from two separate ranks.
Annual costs are estimated for 260 weekdays because demand patterns for these associations on weekends are quite different and
cost models were calibrated for weekday trip counts and profit. Fares would increase ZAR 1 (USD 0.08), or 13.33 percent, for the Higher
evening fare intervention, with costs borne by the passenger rather than the public authority. Because increased fares will likely reduce
demand, passenger counts for this intervention were reduced using an elasticity of −0.40 which results in a 5.33 percent decrease in
demand (Guzman, Gomez & Moncada, 2019; Litman, 2004; Paulley et al., 2006). Drivers will likely still wait for a full van before
departing the rank to maximize their profit. The Headway bonus intervention pays a bonus of ZAR 50 (USD 3.85) per trip only if drivers
depart within 30 min of the previous departure. To accomplish these headways, two vans are needed as cycle times for routes are
25–30 min. Drivers will likely wait until the last minute to maximize the number of passengers, and therefore profit. The public
authority would pay a ZAR 1 (USD 0.08) bonus per transferring passenger under the Transfer bonus scheme. Cost estimates assume that
all 15 passengers transfer because few if any businesses are open in the evening in this area, so all passengers are likely to be transferring
rather than beginning their trip at the PTI. Similar to Higher evening fare, drivers will likely wait for a full van to maximize profit.
The Operating deficit payment intervention has a range because the cost to the public authority depends on the number of
passengers carried per trip. Fewer passengers means a larger shortfall for the authority to cover. The intervention also guarantees
drivers a small ZAR 20 (USD 1.54) profit per trip; with payment only to cover costs, few drivers would be incentivized adequately to
283
C. Plano, et al.
Table 7
Linkages between survey questions and interventions.
Intervention Prefer Fair to Paid fairly Owner can Depart less than Ok if location Use a Tracking Form a Contract with Pax willing to Safety is most
salary pay tax with CFC dictate hours full if pax on shown on phone smartphone device company city pay higher fare important to pax
route
Rank security X
Transfer bonus X X X
284
Headway bonus X X
Operating deficit X X X
payment
E-hailing X X X X X X
Off peak contract X X X X X X X
Higher evening fare X
Transportation Research Part A 132 (2020) 273–289
Table 8
Percentage of drivers willing to accept interventions.
Combined HSS only 7ATA only
Intervention Tax owed Exempt Tax owed Exempt Tax owed Exempt
Rank security 85% 91% 80%

Higher evening fare 52% 69% 39%
Headway bonus 51% 75% 60% 89% 43% 64%
Transfer bonus 39% 58% 49% 71% 32% 48%
Operating deficit payment 39% 58% 49% 71% 32% 48%
Off-peak contract 25% 34% 26% 37% 25% 32%
E-hailing 23% 31% 40% 54% 9% 11%
Notes: When only one value is presented, tax payment is not inherent in intervention.
Table 9
Acceptability and annual cost of interventions for both HSS and 7ATA combined.
Intervention Acceptability Annual cost (ZAR) Annual cost (USD)
Rank security 85%* 108,191 8,322

Higher evening fare 52%* 0 [16,120†] 0 [1,240†]
Headway bonus 51% 299,972 23,075
Transfer bonus 39% 233,672 11,975
Operating deficit payment 39% 225,436 – 284,092 17,341 – 21,853
Off-peak contract 25% 694,095 53,392
E-hailing 23% 124,700 9592
Notes:
*Indicates tax obligation does not apply.
†
Indicates collective cost to passengers.
Acceptability assumes tax obligation.
100% 700,000
90%
600,000
80%
70% 500,000
60% 400,000
50%
40% 300,000
30% 200,000
20%
100,000
10%
0% 0
Rank security Higher evening Headway Transfer bonus Operating Off peak E-hailing
fare bonus deficit contract
payment
Acceptability (%) Annual cost (ZAR)
Fig. 4. Acceptability and annual cost of interventions for both HSS and 7ATA combined. Note: An average value, rather than a range, is shown for
Operating deficit payment.
provide later service. The public authority stops paying at the breakeven point which is estimated at 13 passengers. Drivers are likely
to leave with 12 passengers because they make relatively little profit with 14 or 15 passengers compared to the ZAR 20 (USD 1.54)
guarantee. Therefore, this intervention may have a small effect on headways by encouraging drivers not to wait for a full van but to
leave with 12 passengers instead. There is a risk that guaranteeing profit will encourage drivers to take unnecessary trips for just a
few passengers, however this may be a small risk because the profit guaranteed is less than half of what a driver would make on a trip
under normal circumstances where additional passengers would be picked up along the route (not expected in the evening as all
demand will be outgoing from the PTI to homes).
The Rank security intervention assumes two SAPS officers, one at each rank, at normal salary levels for three hours each night
except for the 12 public holidays when hourly pay doubles per South African law (Department of Labour, 2016). The public authority
currently pays for private security services at the PTI which may provide cost savings if this contract is terminated as SAPS officers are
positioned. However, the authority may also deem it worthwhile to have both private security, which is present overnight, and SAPS
officers that will only be present from 7 pm to 10 pm.
The cost of the contract intervention was obtained by using actual operating cost data from existing MyCiTi feeder routes that
were comparable to the HSS and 7ATA routes on the basis of route distance, speed, and time as well as route characteristics/
285
destinations and surrounding population income. Marginal, rather than total, costs were used because additional routes would
contribute only marginally to fixed costs such as depots and overhead staff. The operating deficit (subsidy cost) was calculated per
month and then for the three hours of evening service to obtain a comparable cost figure.
E-hailing costs include only the cost for app development and maintenance. Estimates for development costs were obtained from
transport technology companies based in Cape Town for an app that would allow for on-demand routing across multiple passenger
pickup and drop-off points as well as in-app payment. This cost was annualized over five years and annual maintenance costs,
assumed to be 17.5 percent of development costs, were added (Chomko, 2012; Patil, 2017). Assuming the app would be made
available only to associations providing feeder services, the annual cost was divided by 11, the number of associations estimated to be
providing feeder services.
4. Discussion
The most acceptable intervention is Rank security, which only requires the addition of a police officer at the rank. The second most
acceptable intervention is one that requires no funding or effort from the public authority, namely Higher evening fare. However, this
does force passengers to pay more for transport, albeit over a service that does not otherwise exist. The risk of this intervention is that
passengers are not willing, or are angered by, the need to pay a higher fare in the evening and this risk may be enough to prevent drivers
from providing service despite the fact that other associations in Cape Town do charge an evening service premium.
Headway bonus is the third most acceptable intervention and ensures 30 min headways in the evening rather than only ad-
dressing the lack of service, an important element for a period with low passenger demand with a service that will otherwise wait for
a full vehicle before departing. This additional service quality does increase the cost to the public authority relative to most others,
but also requires only one technology addition to the vehicles (a tracking device). Without the need for a cashless fare collection
system, passengers can continue to pay in cash, a point of relative satisfaction and importance among passengers of the services
during the day (Behrens et al., 2018).
Transfer bonus and Operating deficit payment demonstrate similar levels of acceptability with increasing costs respectively. Both
Transfer bonus and Operating deficit payment interventions involve a shift to cashless fare collection, which provides some measure
of increased security by making a robbery of the van less attractive to criminals. However, considering that passengers have indicated
that rank security is an important yet dissatisfactory element of the MBT services in this area, and that a survey undertaken by a firm
for the City of Cape Town identified the same need, it may be that SAPS presence may be warranted in addition to these or any of the
interventions (Behrens et al., 2018; Yellowwood, 2018).
The Off-peak contract intervention is almost 2.3 times more expensive than the most expensive of the remaining interventions
(Headway bonus at ZAR 299,972 per year). Both this and the Headway bonus maintain short headways for passengers, though
evening headways across the MyCiTi comparison routes average 38 min while the Headway bonus encourages headways of 30 min.
Whether the increased certainty of improved service quality under a contract is worth the long process of industry transition through
compensation, reformulation into companies, and contract negotiation is an important question outside of the scope of this paper.
Through the Phase 1 MyCiTi transition, operators in the MBT industry did not understand the level of effort required in transitioning
to a fully formal operating environment, causing slow progress despite efforts around capacity-building (Schalekamp, 2017).
The E-hailing intervention could improve security for drivers and passengers by reducing waiting time at the rank and by enabling
door-to-door service, though there is evidence of door-to-door service already being provided by drivers in both associations. No in-
centive is being provided through this intervention, so there is a risk that the app alone will not be enough to encourage service provision
if passenger demand is too low. Route licensing must also be changed from route-based to area-based to allow vehicles to be on-demand
routed based on passenger requests. A shift away from rigid route-based licenses is desired by drivers and owner-drivers who responded
to the survey; 67 percent indicated they could make more money in other places within Mitchells Plain if licenses allowed.
This intervention also highlights issues of equity and access; both drivers and passengers must have smartphones to use the service
and while 65% of drivers across both associations do use a smartphone, this leaves a sizeable group that cannot participate. However,
the public authority could commit to purchasing smartphones and training drivers on how to use them as part of this intervention. By
eliminating this question from the intervention acceptability analysis, the proportion of respondents finding E-hailing acceptable
increases only among 7ATA (an increase of nine percentage points to 18) as well as overall (five percentage points to 28 percent).
While this would make E-hailing more feasible as an intervention, it remains relatively unacceptable to drivers and may not be worth
the additional public authority cost of providing phones and training. More investigation is needed to understand the level of
smartphone usage among passengers or possible alternative technology before this intervention can be realistically considered.
For any of the interventions that require incentive payments, the MBT drivers and owner-drivers must trust the public authority to
pay them especially when payments may only be made weekly or monthly. In the case of incentives involving cashless fare collection
technology, it may be beneficial to include a receipt showing the number of fares paid each evening to allow drivers to have a record
to compare against the city’s payments.
5. Conclusion
These interventions, minus the Off-peak contract, can be described as tweaks to an existing system that have the potential to be
less onerous for both the MBT industry and the public authority. More investigation is needed to understand how these interventions
would work on a detailed level; following this, pilot projects can be used to test them while minimizing risk to the public authority of
a larger-scale failure. Interventions that require much less financial support also are less likely to attract corruption (Timse, 2018).
286
Together this suggests that viable alternatives to paratransit corporatization and contracting exist. Most importantly, these alter-
natives are more affordable and may offer faster quality of service improvements for passengers while building trust between
government and the MBT industry that can pave the way for further reforms in the future.
Overall, this research project sought to provide insight into the MBT operator’s perspective on potential reform paths by using a case
study of two associations. No claim of representativeness is made because of the diversity across associations and across individual business
operations within those associations as each owner/driver is an independent entity. Associations involved in the study have had some
members (owners) attend capacity-building courses taught by University of Cape Town staff and have been involved in the City of Cape
Town’s Taxi Operating Company (TOC) pilot. This may suggest these associations are not representative and findings should not be
generalized. However, average route distance, travel time, and speed of their services do align closely with average characteristics of
feeder/distributor routes city-wide as classified through a recent cluster analysis of MBT route data (du Preez et al., 2019).
Knowing the driver perspective on interventions to encourage off-peak complementarity between scheduled and paratransit PT,
the public authority can begin to tailor reform efforts to those most acceptable and most cost effective. However, there is another half
of this equation: the MBT owners. They must also find interventions acceptable if the public authority expects to be successful. The
second half of this research project will be described in a future paper that brings together these driver results and the results of
engagement with MBT owners on the same interventions. By engaging directly with the MBT industry in Cape Town, the research
team hopes to contribute ‘genuinely co-produced research’ (Schwanen, 2018) to the paratransit reform and integration literature and
enable a more productive working relationship between government and the considerable paratransit industry in South Africa.
Declaration of Competing Interest
None.
Acknowledgements
Thank you to Hazeldene Shuttle Services and Seventh Avenue Taxi Association, specifically Trevor Martin, Mark Tregonning, and
Riedwaan Dollie, for being enthusiastic partners in research. Thank you also to Mareza Valjee and her team at Innovative Transport
Solutions for survey administration. There were a number of experts consulted throughout the research process who kindly assisted:
Andrew Bulman, Babalwa Nyoka, Christo Venter, Elizabeth Hastings, Herrie Schalekamp, Johan Joubert, Nicolette van Niekerk, Paul
Browning, and Stephane Hess. Research was funded by the Volvo Research and Educational Foundations.
Appendix A. Supplementary material
Supplementary data to this article can be found online at https://doi.org/10.1016/j.tra.2019.11.015.
References
Abdel-Aty, M.A., Kitamura, R., Jovanis, P.P., 1997. Using stated preference data for studying the effect of advanced traffic information on drivers’ route choice.
Transport. Res. Part C: Emerg. Technol. 5 (1), 39–50. https://doi.org/10.1016/S0968-090X(96)00023-X.
Aftarobot, 2018, February 23. What you Need to Know About the Launch of Minibus Taxi Hailing App Aftarobot. SME. Retrieved from https://www.smesouthafrica.
co.za/need-know-minibus-taxi-hailing-app-aftarobots-launch/.
Arentze, T., Borgers, A., Timmermans, H., Del Mistro, R., 2003. Transport stated choice responses: effects of task complexity, presentation format and literacy.
Transport. Res. Part E: Log. Transport. Rev. 39 (3), 229–244. https://doi.org/10.1016/S1366-5545(02)00047-9.
Barrett, J., 2003. Organizing in the informal economy: a case study of the minibus taxi industry in South Africa (SEED Working Paper No. 39). Retrieved from
International Labour Organization website: https://ideas.repec.org/p/ilo/ilowps/993581583402676.html.
Behrens, R., Dodgen, S., Fusire, T., Mukhuba, T., 2018. Passenger satisfaction with minibus-taxi feeder services at the Mitchells Plain public transport interchange in
Cape Town. Presented at the Southern African Transport Conference, Pretoria.
Behrens, R., Hawver, H., Birungi, C., Zuidgeest, M., 2017. Case study investigation of unscheduled feeder and scheduled trunk service relationships in Cape Town.
Presented at the Southern African Transport Conference, Pretoria.
Behrens, R., McCormick, D., Mfinanga, D., 2016a. An introduction to paratransit in Sub-Saharan African cities. In: Behrens, R., McCormick, D., Mfinanga, D. (Eds.),
Paratransit in African Cities: Operations, Regulation and Reform. Routledge, New York, pp. 1–26.
Behrens, R., Salazar Ferro, P., 2016. Barriers to comprehensive paratransit replacement. In: Behrens, R., McCormick, D., Mfinanga, D. (Eds.), Paratransit in African
Cities: Operations, Regulation and Reform. Routledge, New York, pp. 199–220.
Behrens, R., Salazar Ferro, P., Golub, A., 2016b. International case studies of hybrid public transport system regulation and complementarity. In: Behrens, R.,
McCormick, D., Mfinanga, D. (Eds.), Paratransit in African Cities. Routledge, New York, pp. 221–243.
Broch, S.W., Strange, N., Jacobsen, J.B., Wilson, K.A., 2013. Farmers’ willingness to provide ecosystem services and effects of their spatial distribution. Ecol. Econ. 92
(Supplement C), 78–86. https://doi.org/10.1016/j.ecolecon.2011.12.017.
Bruun, E., Del Mistro, R., Venter, Y., Mfinanga, D., 2016. The state of public transport systems in three Sub-Saharan African cities. In: Behrens, R., McCormick, D.,
Mfinanga, D. (Eds.), Paratransit in African Cities: Operations, Regulation and Reform. Routledge, New York, pp. 26–56.
Caussade, S., Ortúzar, J. de D., Rizzi, L.I., Hensher, D.A., 2005. Assessing the influence of design dimensions on stated choice experiment estimates. Transportation
Research Part B: Methodological 39 (7), 621–640. https://doi.org/10.1016/j.trb.2004.07.006.
Cervero, R., 1985. Deregulating urban transportation. Cato J. 5, 219.
Cervero, R., 2000. Informal Transport in the Developing World. UN-HABITAT.
Chomko, R., 2012, May 25. Maintaining an app is critical to its overall success. Retrieved October 29, 2018, from FierceWireless website: https://www.fiercewireless.
com/developer/maintaining-app-critical-to-its-overall-success.
Clark, P., Crous, W., 2002. Public transport in metropolitan Cape Town: Past, present and future. Transport Rev. 22 (1), 77–101. https://doi.org/10.1080/
01441640110052092.
Cooke, S., Behrens, R., Zuidgeest, M., 2018. The relationship between transit-oriented development, accessibility and public transport viability in South African cities:
a literature review and problem framing. Presented at the Southern African Transport Conference, Pretoria.
287
De Klerk, A., 2017, January 19. Almost one-third fewer commuters now use Metrorail in Cape Town. BusinessLive. Retrieved from https://www.businesslive.co.za/
bd/companies/transport-and-tourism/2017-01-19-almost-one-third-fewer-commuters-now-use-metrorail-in-cape-town/.
Department of Labour, 2016. Basic Conditions of Employment Act, Amendments of Sectoral Determination 11: Taxi Sector, South Africa. Pub. L. No. 75, vol. 56, no. 1.
du Preez, Zuidgeest, M., Behrens, R., 2019. A quantitative clustering analysis of paratransit route typology and operating attributes in Cape Town. J. Transp. Geogr. 80.
https://doi.org/10.1016/j.jtrangeo.2019.102493.
Espinosa-Goded, M., Barreiro-Hurlé, J., Ruto, E., 2010. What do farmers want from agri-environmental scheme design? A choice experiment approach. J. Agric. Econ.
61 (2), 259–273. https://doi.org/10.1111/j.1477-9552.2010.00244.x.
Gillham, B., 2000. The Research Interview. Continuum, New York.
Golub, A., Balassiano, R., Araújo, A., Ferreira, E., 2009. Regulation of the informal transport sector in Rio de Janeiro, Brazil: welfare impacts and policy analysis.
Transportation 36 (5), 601–616. https://doi.org/10.1007/s11116-009-9215-y.
Guzman, L.A., Gomez, S., Moncada, C.A., 2019. Short run fare elasticities for Bogotá’s BRT system: ridership responses to fare increases. Transportation. https://doi.
org/10.1007/s11116-019-10034-6.
Gwilliam, K., 2008. Bus transport: is there a regulatory cycle? Transport. Res. Part A: Policy Practice 42 (9), 1183–1194. https://doi.org/10.1016/j.tra.2008.05.001.
Hensher, D.A., Rose, J.M., & Greene, W.H., 2015. Applied Choice Analysis, second ed. Retrieved from https://doi.org/10.1017/CBO9781316136232.
Hernández, C.R., Mehndiratta, S.R., 2015. Bus concession contracts and tariff policy: Lessons from the Bogotá and Colombia Experience. In: Presented at the
Transportation Research Board 94th Annual Meeting, Washington, D.C. Retrieved from https://trid.trb.org/view/1337840.
Hitge, G., Vanderschuren, M., 2015. Comparison of travel time between private car and public transport in Cape Town. J. S. Afr. Inst. Civil Eng. 57 (3), 35–43. https://
doi.org/10.17159/2309-8775/2015/V57N3A5.
Johnson, R., Orme, B., 2003. Getting the most from CBC. Retrieved from Sawtooth Software website: https://www.sawtoothsoftware.com/download/techpap/
cbcmost.pdf.
Khosa, M.M., 1992. Routes, ranks and rebels: feuding in the taxi revolution. J. S. Afr. Stud. 18 (1), 232–251. https://doi.org/10.1080/03057079208708312.
Kirby, R.F., Bhatt, K.U., Kemp, M.A., McGillivray, R.G., Wohl, M., 1974. Para-transit: neglected options for urban mobility (No. PB-234320; UMTA-CA-06-0045-74-2).
Retrieved from Urban Institute website: https://www.osti.gov/biblio/7269596.
Kumar, A., & Diou, C., 2010. The Dakar Bus Renewal Scheme: Before and After (SSATP Discussion Paper No. 11). The World Bank.
Lau, S., 1997. Strategies for Improving Jitneys as a Public Transport Mode (Master of Science in Transportation, Massachusetts Institute of Technology). Retrieved from
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.972.2471&rep=rep1&type=pdf.
Litman, T., 2004. Transit price elasticities and cross-elasticities. J. Publ. Transport. 7 (2). https://doi.org/10.5038/2375-0901.7.2.3.
Magubane, N., Manicom, D., 2003, July. Taxi-owners’ perceptions of the government’s recapitalization scheme: a case study of taxi-owners in Pietermaritzburg. In:
Presented at the Southern African Transport Conference, Pretoria. Retrieved from https://repository.up.ac.za/handle/2263/7445.
Mangham, L.J., Hanson, K., McPake, B., 2009. How to do (or not to do) designing a discrete choice experiment for application in a low-income country. Health Policy
Plan. 24 (2), 151–158. https://doi.org/10.1093/heapol/czn047.
McCormick, D., Schalekamp, H., Mfinanga, D., 2016. The nature of paratransit operations. In: Behrens, R., McCormick, D., Mfinanga, D. (Eds.), Paratransit in African
Cities: Operations, Regulation and Reform. Routledge, New York, pp. 59–78.
McFadden, D., 1974. The measurement of urban travel demand. J. Publ. Econ. 3 (4), 303–328.
Mesthrie, U.S., 1993. Tinkering and tampering: a Decade of the group areas act (1950–1960). S. Afr. Hist. J. 28 (1), 177–202. https://doi.org/10.1080/
02582479308671972.
Municipal Finance Management Act. Pub. L. No. 56 of 2003, 26019 Government Gazette, 2004.
National Department of Transport, 2006. Strategy to Accelerate Public Transport Implementation. Pretoria.
National Department of Transport, 2007. Public Transport Strategy. Pretoria.
National Minimum Wage Act. Pub. L. No. 9 of 2018, 2018.
Patil, R., 2017, June 15. What Affects The Cost Of Mobile App Maintenance Services? Retrieved October 29, 2018, from Medium website: https://medium.com/@
ritesh.patil732/what-affects-the-cost-of-mobile-app-maintenance-services-4fa330c9278d.
Paulley, N., Balcombe, R., Mackett, R., Titheridge, H., Preston, J., Wardman, M., White, P., 2006. The demand for public transport: The effects of fares, quality of
service, income and car ownership. Transp. Policy 13 (4), 295–306. https://doi.org/10.1016/j.tranpol.2005.12.004.
Plano, C., Behrens, R., & Zuidgeest, M., 2018. Towards a stated choice methodology to determine minibus-taxi driver willingness to provide off-peak feeder service. In:
Presented at the Southern African Transport Conference, Pretoria. http://hdl.handle.net/2263/69573.
Revelt, D., Train, K., 1998. Mixed logit with repeated choices: households’ choices of appliance efficiency level. Rev. Econ. Stat. 80 (4), 647–657. https://doi.org/10.
1162/003465398557735.
Saddier, S., Johnson, A., 2018. Understanding the operational characteristics of paratransit services in Accra, Ghana: A case study. Presented at the Southern African
Transport Conference, Pretoria.
Saddier, S., Patterson, Z., Johnson, A., Wiseman, N., 2017. Fickle or flexible? Assessing paratransit reliability with smartphones in Accra, Ghana. Transport. Res. Rec.:
J. Transport. Res. Board 2650, 9–17. https://doi.org/10.3141/2650-02.
Salazar Ferro, P., 2015. Paratransit: A Key Element in a Dual System. CODATU, Agence Française de Développement.
Salazar Ferro, P., Behrens, R., 2013. Paratransit and formal public transport complementarity: imperatives, alternatives, and dilemmas. Presented at the World
Conference on Transport Research, Rio de Janeiro.
Salazar Ferro, P., Behrens, R., Wilkinson, P., 2013. Hybrid urban transport systems in developing countries: portents and prospects. Res. Transport. Econ. 39 (1),
121–132. https://doi.org/10.1016/j.retrec.2012.06.004.
Salazar Ferro, P., Muñoz, J.C., Behrens, R., 2015. Trunk and feeder services regulation: lessons from South American case studies. Case Stud. Transp. Policy 3 (2),
264–270. https://doi.org/10.1016/j.cstp.2014.10.002.
Saltzman, A., 1973. Para-transit: taking the mass out of mass transit. Technol. Rev. 75 (8). https://trid.trb.org/view/132882.
Sandoval, E., Hidalgo, D., 2002. TransMilenio: a high capacity – low cost bus rapid transit system developed for Bogotá, Colombia. In: Presented at the Second
International Conference on Urban Public Transportation Systems, Alexandria, VA, https://doi.org/10.1061/40717(148)4.
SARS, 2017. Income Tax: Guide on Income Tax and the Individual (2017/18) (Legal Counsel No. 12).
Schalekamp, H., 2017. Lessons from building paratransit operators’ capacity to be partners in Cape Town’s public transport reform process. Transport. Res. Part A:
Policy Practice 104 (Supplement C), 58–66. https://doi.org/10.1016/j.tra.2017.08.002.
Schalekamp, H., Behrens, R., 2010. Engaging paratransit on public transport reform initiatives in South Africa: a critique of policy and an investigation of appropriate
engagement approaches. Res. Transport. Econ. 29 (1), 371–378. https://doi.org/10.1016/j.retrec.2010.07.047.
Schalekamp, H., Behrens, R., 2013. Engaging the paratransit sector in Cape Town on public transport reform: progress, process and risks. Res. Transport. Econ. 39 (1),
185–190. https://doi.org/10.1016/j.retrec.2012.06.012.
Schalekamp, H., Klopp, J., 2018. Beyond BRT: Innovation in minibus-taxi reform in South African cities. Presented at the Southern African Transport Conference,
Pretoria.
Schalekamp, H., McLachlan, N., 2016. Minibus-taxi operator reforms, engagement, and attitudes in Cape Town. In: Behrens, R., McCormick, D., Mfinanga, D. (Eds.),
Paratransit in African Cities: Operations, Regulation and Reform. Routledge, New York, pp. 174–198.
Schalekamp, H., McLaren, M., Behrens, R., 2017. Exploring cashless fare collection in the context of urban public transport reform in South Africa. Multimodal
Transport Workshop. Presented at the Southern African Transport Conference, Pretoria.
Schulz, N., Breustedt, G., Latacz-Lohmann, U., 2014. Assessing Farmers’ Willingness to Accept “Greening”: Insights from a Discrete Choice Experiment in Germany. J.
Agric. Econ. 65 (1), 26–48. https://doi.org/10.1111/1477-9552.12044.
Schwanen, T., 2018. Towards decolonised knowledge about transport. Palgrave Commun. https://doi.org/10.1057/s41599-018-0130-8.
Scott, A., 2001. Eliciting GPs’ preferences for pecuniary and non-pecuniary job characteristics. J. Health Econ. 20 (3), 329–347. https://doi.org/10.1016/S0167-
288
6296(00)00083-7.
Scott, A., Bond, C., Inch, J., Grant, A., 2007. Preferences of community pharmacists for extended roles in primary care. PharmacoEconomics 25 (9), 783–792. https://
doi.org/10.2165/00019053-200725090-00006.
Siyongwana, P.Q., Binza, M.S., 2012. Challenges facing the transformation of the public transport system in Nelson Mandela Bay, South Africa: history in the making.
J. Contemporary History 37 (1), 191–202.
Smith, L., Hanson, S., 2003. Access to water for the urban poor in Cape Town: where equity meets cost recovery. Urban Studies 40 (8), 1517–1548. https://doi.org/10.
1080/0042098032000094414.
Statistics South Africa, 2018, September 27. Mitchells Plain Main Place [Government]. Retrieved September 27, 2018, from Stats SA website: http://www.statssa.gov.
za/?page_id=4286.
TCT, 2015. MyCiTi Business Plan Update: Phase 1 and N2 Express. Cape Town: City of Cape Town.
TDA, 2014. Operating License Strategy 2013–2018. Cape Town: City of Cape Town.
TDA, 2017. Integrated Public Transport Network Business Plan (No. Technical report V7.1 20170817).
TDA, 2018a. Comprehensive Integrated Transport Plan 2018–2023. Cape Town: City of Cape Town, Transport and Urban Development Authority.
TDA, 2018b. MYFIN 2018 Operational Plan and Strategy for Public Transport 2018–2033. Business Planning Branch, Business Resource Management Department.
Timse, T., 2018, August 30. Billions blown but Ekurhuleni buses remain bogged down. News24. Retrieved from https://www.news24.com/SouthAfrica/News/billions-
blown-but-buses-remain-bogged-down-20180829.
Transportation Research Board, 2013. Transit Capacity and Quality of Service Manual, third ed. (No. TCRP Report 165). Washington, DC.
Ubach, C., Scott, A., French, F., Awramenko, M., Needham, G., 2003. What do hospital consultants value about their jobs? A discrete choice experiment. BMJ 326
(7404), 1432. https://doi.org/10.1136/bmj.326.7404.1432.
Venter, C., Venkatesh, A., 2010. Test and application of stated preference and mixed logit modelling techniques for travel behavior research in a rural context.
Presented at the Southern African Transport Conference, Pretoria.
Von der Heyden, C., Laing, K., Hastings, E., Koch, P., Reddy, T., 2015. Financing IRPTN operations: considerations for cities implementing I(R)PTNs in South Africa.
Retrieved from https://repository.up.ac.za/handle/2263/57784.
Wilkinson, P., 2008. “Formalising” paratransit operations in African cities: constructing a research agenda. In: Presented at the Southern African Transport Conference,
Pretoria. Retrieved from http://repository.up.ac.za/handle/2263/6390.
Wilkinson, P., Mfinanga, D.A., Schalekamp, H., Behrens, R., 2009, April. An international review of paratransit regulation and integration experiences: Lessons for
public transport system rationalisation and improvement in African cities. In: Presented at the 4th international conference on future urban transport: access and
mobility for the cities of tomorrow, Göteborg.
Woolf, S.E., Joubert, J.W., 2013. A people-centred view on paratransit in South Africa. Cities 35, 284–293. https://doi.org/10.1016/j.cities.2013.04.005.
Yellowwood, 2018. PTI User Experience Survey Final Research Presentation – Mitchells Plain PTI. Presented at the City of Cape Town. City of Cape Town.
Glossary
Cashless fare collection: any non-cash system of fare payment including near-field communication (NFC) cards, smartphone apps, Unstructured Supplementary Service
Data (USSD), etc.
Hybrid/hybridity: use of scheduled and unscheduled (paratransit) modes together in an integrated public transport network
Minibus-taxi: South African term for paratransit; typically provided using 15-seat passenger vans
Paratransit: unscheduled public transport provided using small vehicles such as mini- or midibuses via unregistered, cash-based businesses
Public transport interchange: South African term for transfer facility
Rank: minibus-taxi terminal with boarding/alighting as well as holding areas for vehicles, typically with platforms for each route
Taxi association: legally recognized and registered loose affiliation of minibus-taxi owners operating a particular route(s)
289

Transportation Research Part A 132 (2020) 273-289

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

Transportation Research Part A 132 (2020) 273-289

Uploaded by

Copyright:

Available Formats

Transportation Research Part A 132 (2020) 273–289

Contents lists available at ScienceDirect

Transportation Research Part A

Towards evening paratransit services to complement scheduled

ARTICLE INFO ABSTRACT

1.1. South African MBT history

1.2. Cape Town context and reform

2.1. Study area

2.2. Survey method

Fig. 2. Example choice card presented to respondents.

3.1. Choice modelling

structure Status Gender

Commission (7ATA) 469

Target (HSS) 324

All respondents 889

Deﬁnitely no Probably no Unsure Probably yes Deﬁnitely yes

3.2. Driver acceptability of interventions

3.3. Intervention costs

Rank security 85% 91% 80%

Rank security 85%* 108,191 8,322

Acceptability (%) Annual cost (ZAR)

Declaration of Competing Interest

Appendix A. Supplementary material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.tra.2019.11.015.

You might also like