Unlocking Cities Report BCG Uber

Unlocking Cities
The impact of ridesharing in Southeast Asia and beyond

The Boston Consulting Group (BCG) is a global
management consulting firm and the worlds
leading advisor on business strategy. We partner
with clients from the private, public, and not-for-
profit sectors in all regions to identify their
highest-value opportunities, address their most
critical challenges, and transform their enterprises.
Our customized approach combines deep insight
into the dynamics of companies and markets with
close collaboration at all levels of the client
organization. This ensures that our clients achieve
sustainable competitive advantage, build more
capable organizations, and secure lasting results.
Founded in 1963, BCG is a private company with
more than 90 offices in 50 countries. For more
information, please visit bcg.com.
Unlocking Cities
The impact of ridesharing in Southeast Asia and beyond
Commissioned by
Vincent Chin, Mariam Jaafar, Jason Moy, Maria Phong, Shenya Wang,
Matthew McDonnell, and Irfan Prawiradinata
November 2017
ABOUT THIS REPORT
The rise of ridesharing firms such as Uber, Didi, Grab and Lyft over the last eight
years has introduced a new mode of transport to commuters around the world.
Though still nascent in Asia, ridesharing has already begun to influence the
transport landscape. It has the potential to become an important part of the
response to growing transport needs in the region.
Uber has commissioned the Boston Consulting Group to assess the potential
benefits that greater adoption of ridesharing may bring to Asian cities. The
findings in this report were developed through research utilising publicly available
transport data, interviews with transport experts and primary research with
commuters in each city. The cities specifically covered in this report are: Singa-
pore, Kuala Lumpur, Jakarta, Surabaya, Bangkok, Hong Kong, Taipei, Ho Chi Minh
City, Hanoi and Manila.
2 Unlocking Cities
IN BRIEF
Growth in population and wealth Improve vehicle utilisation
have led to an explosion in per KM
transport demand in Asia an
increase of 4x since 1980 Complement public transport
adoption
Despite significant infrastructure
investment, demand has largely Optimise timing of infrastruc-
outstripped supply leading to ture investment
rising congestion and increased
pollution Ridesharing can provide substan-
tial benefits to cities in Asia.
Solutions going forward should However, a combination of
balance between necessary support from both the rideshare
capital investments to expand ecosystem and public sector is
capacity as well as increasing the needed to realise benefits
efficiency of existing assets
Public sector: Review
Ridesharing can play a role in existing regulations to enable
ensuring more efficient use of rideshare ecosystem to deliver
existing assets. This solution higher levels of rideshare
involves three major components: services needed to encourage
(1) Flexible supply base utilising adoption
existing private vehicles, (2)
dynamic routing with smart Rideshare ecosystem:
supply-demand matching, and (3) Enhance rideshare offering,
demand pooling particularly pooling
With these combined compo- Collaboration between

nents, ridesharing can lead to five public sector & rideshare:
major benefits which ultimately Co-developing programs and
reduce congestion and optimise incentives to encourage
infrastructure investment adoption of rideshare and
pooling, particularly in
Support car-light aspirations conjunction with public
and lifestyles transport
Increase occupancy per

vehicle
The Boston Consulting Group 3

Executive Summary
Growth in population and wealth have led to an explosion in transport demand in

Asia increasing 4x across many countries since 1980. While Asian governments
have made significant investments in infrastructure to meet the growing demand,
demand has largely outstripped supply, leading to rising congestion.
Solutions going forward should balance between further capital investments to ex-
pand capacity and initiatives to increase the efficiency of existing assets. Rideshar-
ing is one way to significantly increase the utilisation of existing infrastructure.
Three characteristics of the ridesharing model contribute to its potential as a
cost-efficient part of the overall response to the growing demand for transport in
Asia: (1) Flexible supply base utilising existing private vehicles, (2) dynamic routing
with smart supply-demand matching, and (3) demand pooling.
These three characteristics mean ridesharing offers key benefits for Asian cities, pri-
marily by reducing congestion and optimising public transport investment:
Supporting car-light aspirations and lifestyles: 10-40% of commuters who

plan to purchase a car indicate that they are highly willing to forego purchase if
rideshare matches private car ownership. Purchases will vary by market,
depending on the degree to which these conditions are met along with other
factors such as how cars are viewed as a symbol of wealth. Our analysis suggests
that ridesharing can provide greater access to a car-light lifestyle, resulting in
lower congestion.
Increasing occupancy per vehicle: Pooling of commuters can raise occupancy

per vehicle by an average of 1.7x1 across cities studied, reducing the number of
vehicles needed to meet transport demand and thereby congestion.
Improving vehicle utilisation per KM: Studies show that rideshare may be
more effective in meeting demand when it is at its highest, but reducing supply
when demand is lower. For example, a study in San Francisco found that
rideshare vehicles have approximately half of the miles without passengers
compared to taxis.2
Complementing public transport adoption: Rideshare can support public

transport usage by serving as a first/last mile feeder system. Some studies have
shown higher public transport usage among shared mobility users, suggesting
that ridesharing complements public transport in these cities. Indeed some
cities have programs offering incentives for rideshare use in conjunction with
public transport.
Optimising timing of infrastructure investment: Rideshare has been shown

to better serve outlying areas in some cities, leading governments to partner
with rideshare platforms in order to defer infrastructure investment such as
train stations and parking facilities.
4 Unlocking Cities
Combined, these benefits have the potential to reduce the number of cars on the
road, and tackle congestion levels. Across the cities studied, we estimate ~40%-70%
of private vehicles on the road today could be removed, if rideshare becomes a via-
ble substitute for private vehicle ownership. This will significantly improve conges-
tion in all cases, and almost eliminate it altogether in a number of cities.3
Despite the clear possible benefits from greater rideshare adoption, concerns have
been expressed about the interaction between ridesharing and other transport
modes.The extent and nature of these concerns are typically market-specific it is
important for each city to understand and address them appropriately. To achieve
net positive benefits to Asian cities, several conditions must be met regarding:
Ridesharing substitution of private vehicles: Ridesharing benefits come

from providing greater transport efficiency (people-kilometres) compared to
private vehicles. However cities must ensure substitution of ridesharing for
private vehicles (private cars or motorcycles) and not public transport to provide
net positive benefits to congestion.
Utilisation of taxis: Ridesharing provides new possible avenues of income

generation for drivers. However, to achieve net positive economic benefits, cities
should ensure that taxis can access these technological advancements such as
ridesharing applications, dynamic pricing and smart supply-demand matching
tools to enhance their competitive position. While such access provides one
possible solution to incumbent concerns, the sociopolitical implications of
ridesharing on taxi utilisation will vary by market and must be addressed
appropriately by each city.
Ridesharing must achieve critical mass to realise significant benefits. Rideshare

adoption in Asia today is low (1-6% of KM travelled across cities studied). A combi-
nation of actions by the rideshare ecosystem and the public sector is needed to
drive adoption.
Public sector reform: The ability for rideshare platforms to improve availabili-
ty, timeliness and price is tied to driver supply in each market. Larger supply
pools enable commuters to secure rides more easily, reduce wait times, and
will allow for more attractive pricing. With greater adoption, pooling also
becomes more viable, as more riders create a more efficient pooling net-
work, leading to faster ride times. Supply caps, therefore, potentially limit
the benefits of ridesharing. Restrictions on rideshare apps also prevent
commuters from matching with drivers on demand, effectively curtailing
rideshare usage. Finally, price caps may inhibit a key ridesharing mechanism
which helps match supply with demand in more dynamic fashion during
peak hours. Regulators should take into account the potential benefit of
ridesharing in terms of the overall transport challenge when developing
ridesharing regulations.

Rideshare ecosystem actions: Greater willingness for commuters to adopt
ridesharing, particularly pooling, is essential to realise the full benefits of
ridesharing. Commuters across cities indicate that an improvement in price,
availability and travel times will encourage a higher uptake of pooling. The
rideshare ecosystem must therefore enhance their offering to entice commuters,
which can be facilitated by regulatory support. Rideshare platforms should also
uphold appropriate safety and security measures.
Collaboration between rideshare and public sector: Encouraging rideshare

adoption over less efficient modes of transport (e.g., over private cars, rather
than over public transport) is critical if ridesharing is to create net positive
effects. Programs that incentivize ridesharing in conjunction with public trans-
port systems, such as rider discounts or bundled transport packages, may
encourage this type of adoption.
Ridesharing has the potential to positively contribute to the transport challenge

across Asia. However, substantial gains in adoption are needed in all markets to re-
alise the benefits on a sustained basis. A combination of improved service offerings
from ridesharing platforms as well as support from regulators will be required to
achieve the adoption needed for material benefits.
6 Unlocking Cities
Asias transport growth journey
A sias growth in population and wealth over the last several decades has led
to an explosion in transport demand. Since 1980, the population in East Asia
has increased by nearly 50%, and average wealth, as measured by GDP per capita,
has grown 1.6 times for Asian countries.4 Based on research demonstrating the
relationship between transport demand and population and wealth, we estimate
transport demand has increased, on average, four times per country, since 1980.5
EXHIBIT I: INDEXED ESTIMATED GROWTH IN TRAVEL DEMAND (1980 = 100)

Exhibit I: Indexed estimated growth in travel demand (1980 = 100)
800
Singapore
600 Korea
Malaysia
Vietnam
Thailand
400 Indonesia
Hong Kong
Philippines
Australia
200
Japan
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Source: World Bank; OECD; National Center for Sustainable Transportation; BCG analysis
Source: World Bank; OECD; National Center for Sustainable Transportation; BCG analysis
In response to this growing transport demand, governments in Asia have invested

heavily in transport infrastructure. Asia has driven the majority of infrastructure Transport demand
expenditure over the last decade, with nearly 60% of global infrastructure expendi- has increased, on
ture occurring in the region.6 Transport infrastructure has consequentially im- average, four times
proved significantly, with five Asian countries ranked in the top 15 countries for in- per country, since
frastructure quality today.7 1980
Despite improvements across the region, the current state of transport varies wide-
ly. Asian cities can broadly be classified into three tiers:
Tier 1: Cities with well-developed and formally organised public transport,

covering both road and rail services. Public transport is the predominant
mode of transport for city residents. Within the cities covered in this report,
Singapore, Hong Kong and Taipei fit into this tier.
Tier 2: Cities that have more recently developed extensive, formally

organized public transport networks, covering both road and rail services.
However, in these cities, modes apart from public transport make up the bulk of
transport for city residents. Within the cities covered under this report, Kuala
Lumpur and Bangkok fit into this tier.

Tier 3: These cities have relatively undeveloped public transport networks,
or rely heavily on informal road-based transport networks (e.g. Kopaja in
Jakarta). Out of 140 cities and countries rated by the World Economic Forum,8
the cities classified in this tier tend to be at or below the median percentile
ranking for road and rail infrastructure globally. Within the cities covered in this
report, Jakarta, Ho Chi Minh City, Hanoi, Manila and Surabaya fit into this
tier.
EXHIBIT II: TIERS OF CITIES STUDIED BY ROAD & RAIL INFRASTRUCTURE;
PUBLIC TRANSPORT ADOPTION
Exhibit II: Tiers of cities studied by road & rail infrastructure; public transport adoption
Road and rail infrastructure rating1

7
Tier II:Well developed infrastructure; Tier I:Well developed infrastructure; public transport has the highest
public transport is not the highest share of share of modality
6 modality Singapore Hong Kong
Taipei
Kuala Lumpur
5
Bangkok
4
Jakarta Surabaya
Hanoi
3 Ho Chi Minh
Manila
1
Tier III:Relatively less developed infrastructure
0
0 10 20 30 40 50 60 70 80 90
% of KM travelled in city by public transport

Source:
Source: World
World EconomicForum;
Economic Forum;BCG
BCGAnalysis;
Analysis;Expert
ExpertInterviews
Interviews
Cities in each tier face different sets of transport challenges today. Tier I cities gen-
erally suffer from road congestion during peak travel times, but have low levels of
congestion outside of peak hours. However, public transport in Tier I cities are
showing greater signs of strain resulting in lower levels of public transport satisfac-
tion in some cases, particularly Singapore.9 Looking forward, Tier I cities have aspi-
rations to adopt next-generation modes of transport such as mobility-on-demand
and autonomous vehicles, to improve mobility and to enhance liveability in cities.
Due to the lower adoption of public transport in Tier II cities, road congestion is a
greater challenge. While both Kuala Lumpur and Bangkok have made impressive
gains in developing public transport infrastructure, overall adoption remains rela-
tively low (<25% of KM travelled). The need to further improve overall quality and
ease of access from feeder transport modes have been cited as key barriers to the
desired uptake. Both the Thai and the Malaysian governments have indicated that
maintaining control over vehicle growth and driving uptake of public transport are
transport objectives going forward.
Congestion in Tier III cities is significantly higher than Asian city averages,10 driv-
en by relatively informal and road-based public transport networks, and significant
8 Unlocking Cities
growth in vehicles. Congestion levels, as a result, are high both in peak and non-
peak hours of travel. At current vehicle growth levels, Tier III cities are at risk of
reaching standstill levels of congestion (<10KM/hour) during peak hours by 2022 .
Looking forward, a combination of a significant uptake in public transport as well
as efficient alternatives to vehicle ownership are likely to be needed to curb conges-
tion.
EXHIBIT III: PEAK HOUR CONGESTION (% ADDITIONAL TIME TO TRAVEL IN PEAK HOURS)
Exhibit III: Peak hour congestion (% additional time to travel in peak hours)
150
100
Asia avg
132 134 67%
105 112
50
79
63 65 68 70
57
0
Singapore Hong Surabaya Kuala Taipei Jakarta Bangkok Ho Chi Manila Hanoi
Kong Lumpur Minh City
Vehicle 0.2% 3.4% 6.3% 7.7% -0.8% 10.0% 6.4% 6.4% 4.0% 10.6%
growth (%) 1
1. From 2011-2016 where data available from published government statistics 2. Peak hours dened as 7-9am, 6-8pm
1. From
Note:2011-2016 where
Asia average data
taken available
from from
average published
of East government
Asian cities based onstatistics
TomTom2.trac
Peak index
hours defined as 7-9am, 6-8pm
Note: Asia average
Source: TomTomtaken
tracfrom average
index; GoogleofAPI;
East Asian
Uber; cities based
Government on TomTom
statistics; BCG traffic index
analysis
Source: TomTom traffic index; Google API; Uber; Government statistics; BCG analysis
Ridesharing and its implications for Asian cities

At current vehicle
Ridesharing players such as Uber, Grab and Didi have emerged in Asia over the last growth levels, Tier III
five years, resulting in a new transport options for commuters. While their entry cities are at risk of
has not been without controversy, commuters have benefited from a wider array of reaching standstill
transport options. For the purposes of this report, we will focus on the benefits levels of congestion
ridesharing can potentially provide to a citys transport needs and forward looking during peak hours
ambitions, as well as key challenges which have emerged with the introduction of
this transport model.
We define ridesharing as the combination of several elements:
Flexible supply base: Flexible driver supply base utilising existing private
vehicles, which can scale to meet demand
Dynamic routing and smart matching technology: To efficiently map supply

against demand and route vehicles in a manner that minimizes travel time and
congestion

Demand pooling: Increase vehicle occupancy and thereby passenger-kilometres
delivered per vehicle, based on live demand
In Tier I cities, greater ridesharing adoption could benefit cities by:
Supporting car-light ambitions: In Singapore, Hong Kong and Taipei between

35%-60% of survey respondents indicate some plans to purchase a car in the
next five years (with Taipei being the highest). Of these, ~80% of respondents
state a willingness (with ~20% highly willing) to not purchase a car, should the
availability and timeliness of ridesharing be in-line with car ownership. Ride-
sharing can deliver substantial benefits. At current vehicle utilisation rates,
Should ridesharing should ridesharing replace car ownership, between 40%-60% of cars could be
replace car owner- removed from roads in these cities effectively eliminating congestion during
ship, between 40%- peak hours.
60% of cars could be
removed from roads Reducing congestion: Taxi usage, although still a relatively small portion of
in these cities effec- total modality, is highest in Tier I cities (e.g. ~10% of KM travelled in Taipei).
tively eliminating Because of the ability of ridesharing vehicles to more flexibly match changes in
congestion during transport demand, they can potentially help reduce congestion in peak periods
peak hours without adding to off-peak congestion.
Supporting liveability: With reduced car ownership, land previously used for
parking can be allocated to enhance living conditions, such as additional
housing and social infrastructure.
Tier II cities such as Bangkok and Kuala Lumpur may most benefit from ridesharing
through:
Alternatives to car ownership: Car ownership growth in both Kuala Lumpur

and Bangkok is high, at ~7.7% and 6.4% p.a respectively. Furthermore, commut-
ers in both cities indicate a strong interest in purchasing cars in the next five
years (83% and 88% for Kuala Lumpur and Bangkok respectively). However, over
80% of those who plan to purchase a vehicle indicate that they would consider
not purchasing one, should the availability and timeliness of ridesharing rival
car ownership. Substituting private cars for ridesharing today could eliminate
60% of new cars from the road in these cities going a long way towards
eliminating congestion in peak hours.
Supporting public transport adoption: Despite the significant investment in

quality public transport infrastructure, public transport remains a small portion
of modality (<25% of KM travelled in both cities). Ridesharing has the potential
to act as a feeder to public transport, particularly if applications and incentives
are developed to assist in intermodal transport usage. Outlying areas without
easy access to public transport could most benefit, particularly with sufficient
recruitment of drivers living or working in outlying areas to help serve this
population.
Supplementing incomes: ~25% of drivers in Kuala Lumpur and Bangkok show

a very high interest in becoming rideshare drivers, and an additional 50% are
somewhat willing to consider the role.
10 Unlocking Cities

Tier III cities such as Jakarta, Surabaya, Manila, Hanoi and Ho Chi Minh City may
most benefit from ridesharing through:
Alternatives to car ownership: At current rates of vehicle growth, congestion

may become unmanageable in Tier III cities. In these cities, 80% of commuters
surveyed indicate plans to purchase a car in the next five years. These same
respondents state the highest likelihood, among all cities studied (~40% highly In Tier III cities 80%
willing, 40% willing) to forgo purchasing a vehicle if ridesharing can meet their of commuters sur-
transport requirements on price, timeliness and availability. While the magni- veyed indicate plans
tude of the impact on private car purchase will vary by market depending on to purchase a car in
the degree to which these conditions are met along with other factors such as the next five years.
the extent to which different societies regard car ownership as a symbol of These same respon-
wealth and status, our analysis suggests that ridesharing can provide greater dents state the
access to a car-light lifestyle resulting in lower congestion. highest likelihood,
among all cities
Optimising timing & location of new transport infrastructure: Tier III cities studied to forgo
have ambitious plans to expand transport networks, however roll-out will purchasing a vehicle
require considerable time and funding. Ridesharing can assist governments to if ridesharing can
right-time infrastructure investment, particularly in outlying areas where there meet their transport
may be insufficient demand to warrant fixed asset investment. Ridesharing can requirements
support transport needs in these areas, particularly with sufficient recruitment
of drivers living or working in these areas.
Supplementing incomes: Car owners in Tier III cities indicate higher willing-
ness to drive through a ridesharing platform to increase their incomes. Between
25-33% of car owners indicate very high willingness, and an additional 40-60%
indicate moderate willingness.
Exhibit
EXHIBIT IV: Percentage
IV: PERCENTAGE of respondents
OF RESPONDENTS who
WHO PLAN plan
TO BUY to buy
A CAR a car
WITHIN THEwithin theYEARS
NEXT FIVE next five years
% of respondents
100
88
83 84 83
79 81 79
80
72
64
60 57
51 49
43
40 36
28
21 19 21
20 17 16 17
12
0
Hong Singapore Taipei Kuala Bangkok Hanoi Ho Chi Jakarta Manila Surabaya Average
Tier 1 Tier 2 Tier 3
Mean % yes: 48 86 81
Yes No
Source: BCG survey
Source: BCG survey

Exhibit V: Willingness for a planned car buyer to forego purchase, provided rideshare
meetsEXHIBIT
desiredV: WILLINGNESS
levels of FOR A PLANNED
availability, CARtimeliness
price, BUYER TO FOREGO PURCHASE, PROVIDED RIDESHARE
MEETS DESIRED LEVELS OF AVAILABILITY, PRICE, TIMELINESS
% respondents
100
87 88 89 90
85 85
81 82 82 82
80 9 18
23 28 40 37 42 42 Highly willing
30
45
60
40
73
68
58 54
51 47 51 47 47 Somewhat willing
20 40
0
Singapore Hong Taipei Bangkok Kuala Jakarta Surabaya Manila Hanoi Ho Chi
Source: BCG survey
Source: BCG survey
To give a sense of the magnitude of the potential impact, we assessed a hypotheti-

cal scenario where ridesharing adoption substitutes the most popular private vehi-
cle in each city (car or motorcycle). For example, in Singapore, this would mean
ridesharing becomes 28% of KM travelled, in place of private cars. Under this sce-
nario, ridesharing reduces the number of vehicles required substantially due to the
higher people-kilometres provided by each rideshare vehicle compared to private
cars. Congestion would also decline significantly as a result of the reduction in cars.
Furthermore, significant space could be re-purposed from vehicle parking.
12 Unlocking Cities

Exhibit VI: Average annual
EXHIBIT people-kilometres
VI: AVERAGE travelled per
ANNUAL PEOPLE-KILOMETERS PERvehicle type
VEHICLE TYPE
RIDESHARE
Rideshare VEHICLES
vehicles VS. # 1 PREFERRED
vs. #1 preferred mode of MODE OF PRIVATELY
privately OWNED
owned vehicle VEHICLE
(car (CAR OR MOTORBIKE)
or motorbike)
People KM per vehicle, per annum
80,000
3.4x
1.7x
2.7x
60,000
3.2x
1.3x
1.8x 1.8x
40,000 2.0x
1.9x
1.7x
20,000
0
Hong Kong Singapore Manila Bangkok Taipei Kuala Jakarta Ho Chi Surabaya Hanoi
Lumpur Minh City
Private Motorbike Private Car Rideshare
Source: Government
Source: Governmentstatistics;
statistics;press
presssearch;
search;commuter
commutersurveys;
surveys;BCG
BCGAnalysis
Analysis
ExhibitEXHIBIT
VII: Percentage of private cars, motorcycles and vehicles reduced with rideshare
VII: PERCENTAGE OF PRIVATE VEHICLES AND TOTAL VEHICLES REDUCED WITH RIDESHARE
100%
80%
73%
70% 71%
66%
63% 63%
60%
60% 57% 56% 55%
53%
46% 46%
42%
39% 39%
40% 35%
31%
24%
20%
11%
0%
Singapore Hong Taipei Kuala Bangkok Manila Jakarta Ho Chi Hanoi Surabaya
No. of private
vehicles reduced 0.2 0.3 0.4 2.5 3.7 1 2.5 3.5 2.4 1.0
after rideshare
(million)
Car Motorcycles
% of cars % of total vehicles % of private motorcycle
1.
1.With
With rideshare
rideshare scenario
scenario under which ridesharing
ridesharing replaces
replacesprivate
privatecars
carsasasthe
the#2
#2or
or#3
#3mode
modeofoftransport
transportininrespective
respectivecities
citiesand
and pool
pool
constitutes
constitutes 50%
50% of
of rides 2. Total
Total number
number of vehicles includes
includes private
private cars,
cars, motorcycles,
motorcycles, buses,
buses, taxi
taxi and
and rideshare
ridesharecars,
cars, 3. Total
Total number
number of
cars
cars include
include private cars and ridesharing
ridesharing cars.
cars.
Government statistics;
Source: Government
Source: statistics; BCG
BCG Analysis
Analysis

EXHIBIT VIII: ROAD CONGESTION DURING PEAK HOURS BEFORE VS. AFTER RIDESHARE (2017)
Exhibit VIII: Road congestion during peak hours before vs. after rideshare (2017)
Peak congestion %
150
-88% -85%
8
-85%
-92%
100
-51%
-91% -81%
-90% -72%
-77%
50
0
Singapore Hong Kong Kuala Taipei Jakarta Bangkok Manila Surabaya Ho Chi Hanoi
Lumpur Minh City
After rideshare Before rideshare
Note: Reductions
Note: Reductions in congestion based
in congestion based on
on high-adoption,
high-adoption,high-pooling
high-poolingscenario
scenario
Source: BCG
Source: BCG analysis
analysis
Exhibit IX: Estimated

EXHIBIT space
IX: ESTIMATED SPACEthat can
SAVED BY be savedRIDESHARE
ADOPTING by adopting rideshare
IF RIDESHARE assuming
SUBSTITUTES rideshare
FOR CARS
substitutes for private cars
Hectares saved with rideshare Landmark equivalent
Singapore 872 2x Sentosa
Hong Kong 1,264 67x Victoria Park
Taipei 1,619 197x Botanic Gardens
Kuala Lumpur 9,583 139x Lake Gardens
Bangkok 15,556 273x Lumpini Park
Jakarta 10,647 6x Soekarno-Hatta Airport
Hanoi 339 4x Old Quarter
Ho Chi Minh City 366 17x Zoo 1
Manila 3,362 26x EDSA
Surabaya 545 1x Tanjung Priok
0 10,000 15,000 20,000

Hectares
1.1.
HCMC
HCMCZoo ZooandandGarden
Gardencomplex
complex
Note:Size
Note: Sizeofoflocal
locallandmarks
landmarksvary varygreatly
greatlybetween
between cities.
cities. Area
Arearepresents
representsestimated
estimated total
total flat
at area
area ofof all
all parking
parking lots
lots (existing and needed)
(existing and needed) to
toserve
serveaacity's
cityscar
carpopulation.
population.Area
Areaestimated
estimatedbybyderiving
derivingratio
ratioofofcars
cars(private
(private+ +rideshare)
rideshare) toto estimated
estimated parking
parking lots
lots in in Singapore
Singapore (~2.2)
(~2.2) and
and then
then extrapolating
extrapolating thisthis ratio
ratio to car
to car populations
populations in other
in other markets.
markets. Assumes
Assumes standard
standard parking
parking lot (19m2),
lot (19m2), AreaArea
saved saved
underunder hypothetical
hypothetical
scenario in which rideshare becomes displaces private vehicles in terms of modal split and 50% of
scenario in which rideshare becomes displaces private vehicles in terms of modal split and 50% of rideshare is pooling.rideshare is pooling.
Source:ASEAN
Source: ASEANMaritime
MaritimeWorking
Working Group,
Group,Data.Gov.Sg,
Data.Gov.Sg,FIFA,
FIFA,MapDevelopers/Google
MapDevelopers/Google Maps, Maps, HDB,
HDB,HK HK Census
Census and
and Statistics
Statistics Dept.,
Dept., LTA,
LTA,
Manila
ManilaTimes,
Times,Perdana
PerdanaBotanical
BotanicalGarden,
Garden,URA,
URA,Thanhnien
ThanhnienNews,
News,The TheStraits
StraitsTimes,
Times,Taipei
TaipeiBotanical
BotanicalGarden,
Garden,expert
expertinterviews,
interviews,BCG
BCG analysis
analysis
14 Unlocking Cities

Achieving sustained benefits
On a standalone basis, rideshare can act as a more efficient means of transport
compared to private cars. However, its full benefits may be best realised with a
higher uptake of rideshare and pooling in conjunction with public transport.
Tier I cities are planning significant expansion of public transport. Singapore,

Hong Kong and Taipei have announced planned investments close to $50Bn in
new rail infrastructure by 2022. While this added capacity is sufficient to meet
increased transport demand while maintaining current levels of peak conges-
tion, greater adoption of public transport alone may be insufficient to complete-
ly eliminate congestion during peak hours. Therefore, rideshare could not only
help to alleviate pressure on public transport systems but also serve as a
complementary tool to reduce congestion. We estimate that congestion in Tier I
cities could be cut in half without increasing public transport modality if
rideshare adoption increases to 7-16% across cities.
Tier II cities have recently invested in public transport development, and have
ambitions to significantly increase adoption to reduce congestion. However, we
estimate that even with the planned expansions and full capacity utilisation of
current and future rail lines, both Kuala Lumpur and Bangkok may be unable to
maintain current levels of peak congestion by 2022. Ridesharing, therefore, may
not just be an important mechanism as a feeder to public transport, but will
also be an important way to reduce congestion by substituting against private
car usage. We estimate in 2022 that rideshare adoption of ~10% KM travelled is
needed, alongside utilisation of rail capacity, to maintain congestion levels to
today.
Tier III cities have announced ambitious plans to significantly increase rail-
based public transport capacity. Jakarta, Manila and Ho Chi Minh City have
collectively announced plans to invest over $60Bn in rail infrastructure by 2022.
Despite these ambitious growth plans, we estimate that the added capacity of
rail transport alone will not be sufficient to meet growth in transport demand by
2022. We estimate that ridesharing adoption between 16-40% across these cities
is needed in conjunction with public transport to maintain congestion levels
today.

Exhibit X: Estimated public transport demand in relation to public
EXHIBIT X: ESTIMATED PUBLIC TRANSPORT DEMAND IN RELATION TO PUBLIC
transport capacity TRANSPORT
in 2022 CAPACITY IN 2022
% KM travelled by public transport
100
Tier I Tier II Tier III
80
60
40
20
0
Singapore Hong Taipei Kuala Bangkok Ho Chi Manila Jakarta Surabaya
2017 % KM travelled by public transport
2022 % KM travelled by public transport to maintain current peak congestion
Estimated public transport capacity by 20221
Investment
US$ Billion 20 25 14 6 24 10 51 5 0.09
(2017 2022)
1.Capacity
1.Capacityisisestimated
estimatedbased
basedon oncurrent
currentrail
railnetwork
networkandandnew
newrail
rail lines/existing
lines/existing line
line extensions in operation
operation
before2022
before 2022inineach
eachcity:
city:Thomson
ThomsonEastEastCoast
CoastLine,
Line,Downtown
Downtownlineline33extension
extensionfor
for Singapore;
Singapore; total
total new
new railway
projects
railway equivalent
projects to 25%toof25%
equivalent current capacity
of current in HonginKong;
capacity Hong Circular Line stage
Kong; Circular Line1,stage
Anking Line, Danhai
1, Anking Line, LRT,
WandaLRT,
Danhai LineWanda
stage 1,Line
Xinzhuang
stage 1,Line extension
Xinzhuang Lineforextension
Taipei; MRT Line 2 MRT
for Taipei; for Kuala
LineLumpur; 10 new
2 for Kuala rail lines
Lumpur; 10 and 3
existing
new line extensions
rail lines and 3 existingfor Bangkok; rst Metro
line extensions Line andfirst
for Bangkok; 3 LRT linesLine
Metro for Jakarta; 6 Metro
and 3 LRT lines Rails (total 109
for Jakarta; 6 KM)
for HoRails
Metro Chi Minh; 6 new
(total 109 KM) railway
for Holines
Chi (total
Minh;246
6 newKM) for Manila;
railway one monorail
lines (total 246 KM)forforSurabaya
Manila; one monorail for
Surabaya
These benefits previously discussed are supported by recent studies conducted on

ridesharing, shared mobility and car-pooling around the world. These studies sug-
gest the following benefits:
Benefit 1: Supporting a car-light lifestyle

Shared mobility has been shown to suppport car light lifestyles in some cities. For
In London, research- example, in London, researchers surveyed car-share users and found that 31% of us-
ers surveyed shared ers declined to purchase a car they otherwise would have purchased while 6%
mobility users and of car owners planned to or had recently disposed of a car due to ridesharing
found that 31% of availability.11 By comparison, in Austin, Texas, researchers found that when Uber
users declined to and Lyft were temporarily suspended in that city, roughly 40% of those affected
purchase a car they switched to a personal vehicle as their primary transport mode and approxi-
otherwise would have mately 9% purchased a vehicle in response to the suspension. 12, 13
purchased
16 Unlocking Cities

Benefit 2: More passengers per vehicle
A key way ridesharing can reduce congestion is via increased vehicle occupancy.
This benefit was demonstrated in Jakarta where, in 1992, the government intro-
duced a policy that required vehicles to carry at least three occupants when travel-
ling on main routes during peak hours (3-in-1 policy). This policy was lifted in 2016
due, at least in part, to concerns regarding the informal passenger-for-hire (i.e. jock-
ey) economy that emerged as a result.14
A recent study by researchers at Harvard and MIT universities found that following
the repeal of this policy, morning and evening congestion on the newly-liberalized Following the repeal
routes leaped by a staggering 46% and 87%, respectively. Moreover, not only did of the 3-in-1 policy,
congestion jump on those central Jakarta roads where car-pooling was previously morning and evening
mandated, it increased in areas that were never subject to the pooling rule in congestion on the
the first place. In the hour following the evening peak, for example 19:00-20:00, newly-liberalized
the repeal of this policy coincided with a roughly 50% increase in delays.15 routes leaped by a
staggering 46% and
While this example is not ridesharing-specific, it illustrates the benefits carpooling 87%, respectively
facilitated by ridesharing. Furthermore, there is evidence that pooled rides can
become a substantial portion of ridesharing trips. Lyft, for example, reported in
2015 that the companys pooled offering represented 50% of total Lyft trips in San
Francisco, and 30% of total Lyft trips in New York City. In Southeast Asia, Ubers
pooled option represented approximately 25% of total trips in August 2017.16
Benefit 3: Greater vehicle utilisation per kilometre

A common challenge in cities is matching transport supply with demand to ensure
sufficient supply during peak times, but reducing supply during off-peak time, to
minimize KM travelled without passengers (unproductive miles). Research in San
Francisco indicated rideshare vehicle unproductive miles is approximately
half of taxis (as a percentage of total miles).17 Because ridesharing can respond
more flexibly to demand, ridesharing vehicles are potentially more efficient in
meeting demand, without adding to congestion when there is lower demand.
Benefit 4: Complementing public transport to accelerate adop-

tion
Studies have also shown that in addition to reducing car ownership, shared mobility
users are more likely to increase their use of public transport. A study published by
the National Academy of Sciences, which covered several major US cities, found
that 43% of shared-mobility users reported an increase in their use of public trans-
port, while only 28% of individuals reported using public transport less.18 Where
public transport use has increased, the study suggests that ridesharing is used to
complement public transport, and can support a car light lifestyle.
Transit authorities are recognizing the potential for ridesharing to act as a feeder
mechanism to public transport. In Portland, Oregon, for example, a local transit au-
thority (TriMet) has integrated rideshare booking capabilities into its public transit
app, as a means to enhance intermodal efficiency to public transport hubs.

Benefit 5: Helping optimise infrastructure timing & location
Another benefit of existing rideshare models is improved transport coverage of ar-
eas outside of the core metropolitan space. Studies in Manhattan found that outly-
ing areas were generally better served by rideshare, compared with taxis.19
Ridesharing, therefore, supports transport needs where there is less access to public
Some municipalities transport, and can serve as a bridging mechanism for infrastructure development
have explored ride- in outlying areas. In the US, for example, some municipalities have explored ride-
sharing as an alterna- sharing as an alternative to infrastructure investment. In 2016, a New Jersey suburb
tive to infrastructure subsidized Uber rides to the local public transit hub, instead of using the funds to
investment expand parking. Numerous other transit departments have struck deals with ride-
share platforms to provide transport services to otherwise underserved areas.
Despite the clear possible benefits of ridesharing, concerns have emerged about the
interaction between ridesharing and other transport modes such as taxi operators
and public transport players. BCG has therefore explored these concerns and poten-
tial ways forward. From our assessment, we found that a net positive outcome can
be realised for all stakeholders ridesharing is not and need not be a zero sum
game.To achieve net positive benefits to Asian cities, several conditions must be
achieved regarding:
Ridesharing substituting against private vehicles

Ridesharing benefits are obtained by providing greater transport efficiency (peo-
ple-kilometres) compared to private vehicles. However, to provide net positive ben-
efits for congestion, cities must ensure substitution of ridesharing for private vehi-
cles (private cars or motorcycles) and not public transport. While there is evidence
that rideshare can supplement public transport and support car-light lifestyles (see
above), there is mixed evidence suggestingthat ridesharing may substitute for pub-
lic transport use under certain conditions.20
This challenge is potentially most significant for Tier I cities in Asia that currently
rely heavily on public transport. However, among the Tier I cities studied, the price
differential between private vehicle ownership and public transport is large given
government control over vehicle prices. Therefore, assuming rideshare prices re-
main more attractive in comparison to car ownership than public transport, the risk
of public transport substitution may not be significant.
This risk can be further mitigated by rideshare platforms and governments working
together to establish programs that make ridesharing services an appealing comple-
ment to public transport. For example, governments can work with ridesharing
platforms to provide commuters with live inter-modal travel data and to establish
discounts or pooling schemes for feeder transport to arterial public transport infra-
structure.
Utilisation of taxis
The rise of rideshare has been perceived to reduce taxi ridership in some cities. For
example, data from the Land Transport Authority of Singapore suggests that the
proportion of taxis sitting idle in yards has increased from 2016 to 2017.21 However,
the Ministry of Transport in Singapore has also suggested that rideshare has served
as a positive complement to taxis, particularly in peak hours.22, 23 In addition, the
emergence of rideshare technologies may have encouraged taxis to adopt more so-
18 Unlocking Cities

phisticated technological advancements such as electronic applications, dynamic
pricing and smart supply-demand matching tools enhancing their competitive po-
sition and ultimately benefitting commuters. Furthermore, in Sydney, taxi ridership
has grown since the entrance of rideshare, suggesting that the risk of disruption to
taxis is uncertain and market specific.
Governments can also play a role in ensuring taxi companies improve their compet-
itive position while offering commuters better outcomes.For example,taxis should
be able to access the same technologies available to ridesharing vehicles. Both taxis
and private vehicles can form part of the flexible supply base necessary to realise
the congestion benefits outlined above. In particular, governments should ensure
that taxis can use apps to connect with passengers, and ensure that taxis can avail
themselves of supply-demand matching mechanisms such as dynamic pricing.
Partnerships between rideshare platforms and taxi companies can also benefit taxi
drivers. Recent examples of partnerships between rideshare platforms and taxi
companies include UberTAXI in Taiwan, UberFLASH in Malaysia and Grabs part-
nerships with multiple Singaporean and Vietnamese taxi companies. These part-
nerships promise to benefit taxi drivers by offering them access to technology
which may allow more responsive matching of supply to demand, thereby increas-
ing vehicle utilisation and ridership. These partnerships also benefit drivers by of-
fering access to large networks of potential passengers.
We believe net positive outcomes can be realised across stakeholders in the trans-
port landscape. Demand for transport will continue to grow across Asian cities,
leading to opportunities for incumbent transport models to evolve and for new
transport models to enter ultimately leading to better transport outcomes for
commuters.
The path forward: realizing the benefits of ridesharing
Ridesharing has the potential to support the growth in transport needs across Asian
cities in a more sustainable and more efficient manner than private car ownership.
However, both the rideshare ecosystem and public sector must play a role to realise
this outcome.
Support needed from public sector
Regulatory restrictions for ridesharing vary widely across Asia with some cities
adopting a more open market position, and others imposing explicit restrictions.
While these regulations reflect differing circumstances across cities, more open
stances may be needed if ridesharing platforms are to achieve the improved levels
of service that will encourage higher adoption:
Barriers to application usage: Mobile applications are essential to

enable ridesharing platforms to dynamically match supply with demand
and ensure commuters can connect with these services on request.
Restrictions on the use of applications thus significantly limit the benefits
ridesharing can bring to markets.

Vehicle supply restrictions: Placing limits on rideshare may inhibit
adoption as ridesharing platforms may be unable to meet the levels of
availability and timeliness that will encourage adoption and drive substi-
tution against private vehicles. For example, in Hong Kong, the stated
supply caps of 1,500 private hire cars24 is approximately 70x lower than
In Hong Kong, the the estimated number needed to reduce congestion by half while main-
stated supply caps of taining current levels of public transport adoption. In Ho Chi Minh City,
1,500 private hire cars where the government has announced plans to limit taxis and contract
is approximately 70x cars to 12,700 by 2020, the cap (assuming the limit is taken entirely by
lower than the contract cars) is 170x too low to accommodate the number of rideshare
estimated number vehicles needed to achieve the benefits articulated above.
needed to reduce
congestion by half Price controls: Price is an important lever in the ridesharing model. It
while maintaining encourages flexible driver supply to meet peaks in demand, which in turn
current levels of impacts rideshare availability and travel time. Price also helps manage
public transport commuter demand encouraging, where possible, commuters to travel
adoption outside of peak demand periods. While price caps were historically
instituted for commuter-protection in certain traditional transport mod-
els, they also inhibit a key ridesharing mechanism which helps match
supply with demand in more dynamic fashion.
Actions needed from ridesharing ecosystems
Greater willingness for commuters to adopt ridesharing, specifically pooling,

is essential to achieve the benefits of ridesharing. Enhanced price, availabili-
ty and more attractive travel times are needed across cities to encourage
adoption.
Price: Across cities surveyed, approximately 80% of respondents who do

not currently use pooling cite prices as a key reason. The majority of
respondents indicate that prices must be at least 25% cheaper than their
current preferred mode of transport to adopt pooling.
Availability: Rideshare availability is critical to higher adoption, as lower

availability can translate to longer wait times or difficulty securing transport
when needed. Lower than desired availability is cited by between 60-80% of
respondents as a key driver for not adopting pooling. Across cities, respondents
note that pooling services must be at least as easily available as their current
preferred mode of transport to adopt pooling.
Travel time: Longer commute times due to the nature of pooling appear to be
more of an issue in cities like Jakarta, Kuala Lumpur, Bangkok and Ho Chi Minh,
where respondents indicate the travel time would be substantially higher than
their current mode of transport.
20 Unlocking Cities

Consequently, raising ridesharing service levels are essential to encourage greater
adoption. While ridesharing platforms must therefore enhance their pooling prod-
ucts, improved regulatory conditions can also help rideshare platforms achieve
their desired outcomes. Rideshare platforms should also uphold appropriate safety
and security measures.
Collaboration needed between governments and rideshare ecosystem
Greater collaboration between government agencies and ridesharing platforms is

needed to encourage modality shifts from less efficient modes of transport (e.g. pri- Greater collaboration
vate cars) to ridesharing . This substitution is essential to create net-positive bene- between government
fits to congestion. Such collaboration is showing promise in a number of US cities. agencies and ride-
For example, Uber has partnered with the transit authorities in Atlanta, Los Ange- sharing platforms is
les and Minneapolis to provide a discount to commuters using Uber to complement needed to encourage
public transport. Programs such as guaranteed ride home in Washington DC offer modality shifts from
commuters who regularly use pooling (twice a week) reimbursement for emergency less efficient modes
travel outside of peak hours. As both ridesharing and public transport service levels of transport (e.g.
improve, such collaboration can provide important incentives to commuters to private cars) to
adopt ridesharing in conjunction with public transport and to maximize the trans- ridesharing
port benefits of both networks.
Ridesharing has the potential to positively impact the transport environment across
Asia. While the existing benefits for ridesharing and pooling vary, substantial
growth in adoption is needed in all markets to realise benefits on a sustained basis.
A combination of improved service offerings from ridesharing platforms as well as
support from regulators will be required to achieve the adoption needed for materi-
al benefits.

Notes
1. Based on average occupancy for private cars ranging from 1.6 to 2.8 across Asian cities surveyed
2. San Francisco County Transportation Authority, 2017
3. Estimated based on substitution of private cars for rideshare vehicles (50% pool). Cities that would
nearly achieve speed-limit travel times under this scenario are Singapore, Hong Kong, Kuala Lumpur
and Taipei
4. For countries in Asia among top 100 largest economies by GDP in 2016
5. Based on indexed population and GDP per capita (constant) growth from 1980-2016 for Asian
countries among top 100 GDPs in world
6. Oxford Economics Global Infrastructure Outlook ( July 2017)
7. World Economic Forum Global Competitiveness Report (2016)
8. World Economic Forum Global Competitiveness Report (2016)
9. Based on BCG survey among commuters with ~300 respondents per city
10. Based on 2017 traffic data from TomTom, Google traffic and Uber
11. Le Vine & Polak, 2017
12. Hampshire, Simek, Fabusuyi, Di, & Chen, 2017.
13. During this period, other rideshare platforms continued to operate within Austin. The portion of
former Uber and Lyft customers who migrated to this platforms is roughly the same as the portion
that migrated to cars (roughly 40%).
14. Anggun Wijaya, 2016; Tempo.co, 2016
15. Hanna, Kreindler, & Olken, 2017
16. Lyft Blog,2015; Uber Data, 2017
17. San Francisco County Transportation Authority and Northeastern University
18. Shared mobility and the transformation of public transit, Feigon, Murphy, 2016. Shared mobility
defined as public transit, bike sharing, car sharing, ridesharing, and similar modes
19. San Francisco County Transportaton Authority, 2017; Schaller, 2017
20 For example, in the US, according to the National Academy of Sciences study of shared mobility
users referenced earlier, 43% of individuals reported an increase in their use of public transport, while
only 28% of individuals reported using public transport less. In addition. However, a study from UC
Davis Disruptive Transportation: The Adoption, Utilization, and Impacts of Ride-Hailing in the United
States (October 2017)suggest that ridesharing may decrease use of bus and light-rail services by 6%
and 3% respectively in several major metropolitan US cities.
21. The proportion of Singapore taxis sitting idle in yards increased roughly 80% in the first five
months of 2017 over the same period in 2016
22. Straits Times Feb 2017, Abdullah; 967,000 taxi trips daily in 2013 to 954,000 trips daily last
year (2016)
23. Dont stop taxi industry from adapting to competition: Ng Chee Meng; Channel News Asia
April 2017
24. Road Traffic (Public Service Vehicles) Regulations,regulation 19(1)) for private hire car service
22 Unlocking Cities

About the Authors
Vincent Chin, is a Senior Partner based in Singapore and is the Global leader of BCGs Public Sec-
tor practice. He brings 20+ years of experience in working with governments and policy makers
globally. He can be contacted via email at chin.vincent@bcg.com.
Mariam Jaafar, is a Partner in our Singapore office and a member of Singapores Committee on
the Future Economy. She is also on the Board of GovTech, the agency responsible for implementa-
tion of Singapores Smart Nation agenda. She has worked closely with multiple public sector cli-
ents across APAC on topics related to the digital economy, giving her a unique understanding of the
policy perspectives of the Singapore government. She can be contacted via email at jaafar.mari-
am@bcg.com.
Jason Moy, is a Principal based in Singapore and co-leads BCG Vietnam. He has ~15 years of expe-
rience in working with consumer businesses across APAC, Europe, and the US. He has also sup-
ported various SEA government agencies to successfully implement public policy initiatives (e.g.,
labour productivity) across industries. He can be contacted via email at moy.jason@bcg.com.
Maria Phong, is a Principal in the Singapore office. She has over 7 years of experience consulting
with Asian companies and governments on strategic growth and logistic topics such as smart trans-
portation, transport megaprojects and labour productivity. She can be contacted via email at
phong.maria@bcg.com.
Matthew McDonnell is a Consultant in the Jakarta office. Shenya Wang and Irfan Prawiradi-
nata are Associates in the Singapore and Jakarta offices respectively. They can be contacted via
email at mcdonnell.matthew@bcg.com, wang.shenya@bcg.com and prawiradinata.irfan@bcg.com.
Acknowledgements
The authors would like to thank Panagiota Papakosta and the broader BCG GAMMA team for their
traffic congestion analysis, Kirsten Lees for editorial support, Kim Friedman and Varvara Egorova
for design and production assistance, the BCG knowledge teams and Visual Services.
For Further Contact

If you would like to discuss this report, please contact one of the authors

Appendix: Detailed methodology
The focus of this report is to assess the potential benefit of ridesharing on key
Asian cities. We conducted both qualitative and quantitative analysis to develop the
findings in this report.
Analysis conducted
1. Qualitative research:
The qualitative research conducted as part of this report takes two primary
forms:
a. BCG survey of commuter sentiments in cities: The objective of these
surveys was to develop an understanding of commuter satisfaction with
existing transportation options, their reasons for using or not using ride-
share and pooling, their likelihood to adopt these methods, the impact of
ridesharing on car ownership and commuter desire to become rideshare
drivers.
b. Literature review of recent ridesharing studies covering the benefits and

key conditions which must exist for cities to achieve net positive benefits.
2. Quantitative research:
The quantitative research conducted as part of this report was used to model
the potential benefits provided by ridesharing and pooling under different
adoption scenarios.
1a. BCG survey of commuter sentiments in cities
Survey methodology
The BCG survey, conducted in September-October 2017, covered approximately 300
commuters per city. Commuters surveyed ranged across all types of transportation.
The surveys covered the following cities: Hong Kong, Singapore, Taipei, Kuala Lum-
pur, Bangkok, Ho Chi Minh City, Jakarta, Manila, Hanoi and Surabaya.
24 Unlocking Cities

Key survey findings
Finding 1: Rideshare and pooling are relatively small proportion of modality
BCG survey results suggest that, on average, rideshare represents approximately

10% of transport taken. Of these trips, pooling makes up ~20% of rideshare trips.
These results support our quantitative assessment that rideshare adoption and
pooling is nascent in Asia.
Appendix exhibit I:APPENDIX EXHIBIT I: PERCENTAGE OF RESPONDENTS WHO USE RIDESHARE

Percentage of respondents who use rideshare
% total transportation mix
50
40
30
20 19
13
11 11 11 12
10
10 Mean 10
7
5
3
0
Hong Kong Taipei Singapore Bangkok Kuala Hanoi Ho Chi Jakarta Surabaya Manila
Lumpur Minh City
Source: BCG survey
Source: BCG survey

Among rideshare users, pooling on average makes up 20% of trips.
APPENDIX EXHIBIT II: PERCENTAGE OF RIDESHARE RESPONDENTS WHO USE POOLING

Appendix exhibit II: Percentage of rideshare respondents who use pooling
% total trips, weighted avg.
50 Tier I Tier II Tier III
40
30
25
24
23 22
20 21
19 Mean 21
20 17 17 18
10
0
Hong Kong Singapore Taipei Kuala Bangkok Hanoi Surabaya Jakarta Manila Ho Chi
Lumpur Minh City
Source: BCG survey
Source: BCG survey

Finding 2: Commuters cite price as the strongest factor inhibiting rideshare adop-
tion
We surveyed current non-users of rideshare for the key reasons they do not use this
mode of transport. In all but two of the cities surveyed, respondents cited price as
being higher than their current primary mode of transport, as the strongest reason.
At the other end of the spectrum, awareness of rideshare offerings was seen as the
least severe inhibitor to rideshare adoption across all the cities studied.
Appendix exhibit
APPENDIX EXHIBITIII: Reasons
III: REASONS cited
CITED FORfor
NOTnot adopting
ADOPTING rideshare
RIDESHARE in comparison
IN COMPARISON to respon-
TO RESPONDENT
PREFERRED MODE OF TRANSPORT
dent preferred mode of transport
Higher Greater Lower Less Driver
price travel time availability awareness safety
Hong Kong
Singapore
Taipei
Kuala Lumpur
Bangkok
Hanoi
Ho Chi Minh City
Jakarta
Manila
Surabaya
All cities
average
Legend
1 2 3 4 5
Strongest Weakest
agreement agreement
Note: Reasons for non-adoption are relative to rideshare non-users' preferred modes of transport
Note: Reasons for non-adoption
Source: BCG survey are relative to rideshare non-users preferred modes of transport
Source: BCG survey
26 Unlocking Cities

Finding 3: Majority of respondents state that improvements in price, availability
and travel time relative to their current primary mode of transport, are needed for
them to adopt pooling
Understanding the circumstances under which commuters will be willing to adopt
15 ridesharing, particularly pooling, is important to drive increased adoption. In the

majority of cities surveyed, current non-users state they would be willing to adopt
pooling, should prices become 25% cheaper than their current mode of transport.
Appendix exhibit IV: % current non-pool users who would adopt pooling if:
Price is up to APPENDIX
25%PRICE
EXHIBIT IV: % CURRENT NON-POOL USERS WHO WOULD ADOPT POOLING IF:
cheaper
IS UP TOthan currentTHAN
25% CHEAPER primary mode
CURRENT of transport
PRIMARY MODE OF TRANSPORT
%
80
60
50
40 74
66
65
66 56
53
48 50
20 41 39
0
Singapore Taipei Hong Kong Kuala Bangkok Jakarta Manila Surabaya Hanoi Ho Chi
Lumpur Minh City
Cities where >50% of respondents willing to adopt pooling under stated condition
Source: BCG commuter survey; BCG analysis


In addition, slight improvements to availability vs. their current primary mode of
transport would be sufficient to drive adoption of pooling in the majority of cities
surveyed.
APPENDIXV:
Appendix POOLING
exhibit EXHIBIT V: % CURRENT
% current NON-POOL
non-pool USERS
users whoWHO WOULD adopt
would ADOPT POOLING
poolingIF: if:
IS SLIGHTLY MORE EASILY AVAILABLE THAN CURRENT MAIN MODE OF TRANSPORT
Pooling is slightly more easily available than current main mode of transport
%
60
50
40 71
69
60
52 54 51
47 51 52
43
20
0
Singapore Hong Kong Taipei Kuala Bangkok Surabaya Jakarta Manila Ho Chi Hanoi
Lumpur Minh City
17 Source: BCG commuter survey; BCG analysis

Finally, the majority of respondents indicated slight improvement to travel speed

vs. their current primary mode would be sufficient for to utilise pooling.
APPENDIX EXHIBIT VI: % CURRENT NON-POOL USERS WHO WOULD ADOPT POOLING IF:
APPENDIX EXHIBIT VI: IS%SLIGHTLY
POOLING currentFASTER
non-pool users who
THAN CURRENT MAINwould adopt
MODE OF pooling if:
TRANSPORT
Pooling is slightly faster than current main mode of transport
%
60
50
40 74
65
54 58 57 58
50 50
44 47
20
0
Singapore Taipei Hong Kong Kuala Bangkok Surabaya Manila Jakarta Ho Chi Hanoi
Lumpur Minh City
28 Unlocking Cities

These findings are encouraging as the majority of respondents indicate that rela-
tively small adjustments to pooling service levels could significantly increase adop-
tion.
Key Finding 4: Commuters show high willingness to decrease car ownership if

desired rideshare service levels can be met
BCG surveyed the stated likelihood of commuters to purchase a car in the next 5
years. Respondents in Tier II and Tier III cities show very high propensity to pur-
chase a car.
Appendix exhibit
APPENDIXVII: Percentage
EXHIBIT of respondents
VII: PERCENTAGE who
OF RESPONDENTS plan
WHO PLANto
TObuy
BUY a car WITHIN
A CAR withinTHE
the next five
NEXT
FIVE YEARS
years
% of respondents
100 Yes No
88
84 83
83 81 79
79
80
72
64
60 57
51 49
43
40 36
28
21 21
17 19 17
20 16
12
0
Hong Singapore Taipei Kuala Bangkok Hanoi Ho Chi Jakarta Manila Surabaya Average
Mean % yes: 48 86 81
Source: BCG survey

Source: BCG survey

We also surveyed the stated willingness to forego purchasing a car in the event
ridesharing achieves their desired level of service. The results indicate that, on
average, 80% of respondents who previously indicated plans to purchase a car were
either highly willing or somewhat willing to not purchase a car.
APPENDIX EXHIBIT VIII: WILLINGNESS FOR A PLANNED CAR BUYER TO FOREGO PURCHASE, PROVIDED
Appendix exhibit VIII: Willingness
RIDESHARE for a planned
MEETS DESIRED LEVELS OF car buyer to
AVAILABILITY, forego
PRICE, purchase, provided
TIMELINESS
rideshare meets desired levels of availability, price, timeliness
% respondents
100
87 88 89 90
85 85
81 82 82 82
80 9 18
23 37
30 28 40 42 42 Highly willing
45
60
40 73
68
58 54
51
47 51 47 47
20 40 Somewhat willing
0
Singapore Hong Taipei Bangkok Kuala Jakarta Surabaya Manila Hanoi Ho Chi
Source: BCG survey

Source: BCG survey
Receptivity and enthusiasm for such a scenario varied somewhat with the citys
current level of transport infrastructure development and public transit adoption.
General receptivity to forgoing a planned car purchase (somewhat + highly agree)
varied modestly between city tiers, at 83%, 82% and 88% for Tier 1, Tier 2 and
Tier 3 cities, respectively. More pronounced among the cities was enthusiasm for
the idea, with only 16% of respondents in Tier 1 cities saying they highly agree,
compared to 29% and 41% for Tier 2 and Tier 3 cities, respectively.
30 Unlocking Cities

Key Finding 5: Majority of drivers somewhat willing to consider driving for ride-
share to supplement income
While the aforementioned topics primarily concern consumer behaviour as it re-

lates to rideshare, of equal import is the supply of individuals willing to work as
rideshare drivers and the incentives that drive that behaviour.
APPENDIX EXHIBIT
Appendix exhibit IX: SURVEY
IX: Survey RESPONDENTS
respondents STATEDwillingness
stated WILLINGNESS TO
toDRIVE
driveFOR RIDESHARE
for rideshare
% respondents
100
89 90
86
81
80 75 77 77 76
72 32 31 Highly
7 9 30
24 willing
24
60 27 31
53
8
40
69 67
57 56 58 58 Somewhat
53
45 44 45 willing
20
0
Taipei Hong Kong Singapore Bangkok Kuala Jakarta Surabaya Manila Ho Chi Hanoi
Lumpur Minh
Source: BCGBCG
Source: survey
survey
While car owners across the markets studied were generally positive about the
prospect of working as a rideshare driver, the general receptivity and enthusiasm
for doing so was more varied, relative to consumers general willingness to adopt
rideshare. Overall across the cities studied, 76% either somewhat or highly agreed
with the statement that they would be willing to use their own cars to work as a
rideshare driver. Enthusiasm for driving was lower than the consumer metrics ex-
amined previously, with only 8% of Tier 1 car owners expressing strong agreement
and 24% and 30% of Tier 2 and Tier 3 respondents strongly agreeing, respectively.

1b. Literature Review
Due to the relatively nascent nature of ridesharing in Asia, BCG reviewed studies
which assessed the impact of ridesharing in markets where ridesharing is more
prominent and commands a larger share of modality. Our review of this literature
surfaced both benefits and key conditions which must be met to achieve net posi-
tive benefits:
Benefit 1: Curbing vehicle growth
In various U.S. cities, research found that average number of cars per household
was roughly a third less in car share, rideshare, and bike share households vs.
households that did not use those shared mobility options. (Feigon & Murphy, 2016)
In Austin, Texas, researchers found that a when Uber and Lyft were temporarily
suspended in that city, roughly 40% of those affected switched to a personal ve-
hicle as their primary transport mode and approximately 9% purchased a ve-
hicle in response to the suspension. 1, 2 (Hampshire, Simek, Fabusuyi, Di, & Chen,
2017).
Benefit 2: More passengers per vehicle
In 1992, the Jakarta government introduced a policy where vehicles were required
to carry at least three occupants when travelling on main routes during peak hours,
a policy known locally as 3-in-1. This restriction was in excess of the more common
high occupancy vehicle (HOV) standard of +2, in part because many private car
owners also hire a driver. (Hanna, Kreindler, & Olken, 2017) In Mid-2016, the policy
was then scrapped, first temporarily, then permanently, due at least in part to con-
cerns regarding the informal passenger-for-hire (i.e. jockey) economy that grew in
response to those trying to circumvent the restrictions. (Anggun Wijaya, 2016)
(Tempo.co, 2016)
Whatever the reasons for its demise, the congestion effects of eliminating Jakartas
HOV policies were staggering. A recent study by researchers at Harvard and
MIT universities found that following 3-in-1s elimination, morning and eve-
ning congestion on the newly liberalized routes leaped by a staggering 46%
and 87%, respectively. In some cases, average speeds slowed to roughly 11km/hr.
hardly more than 2x average walking speed. (Hanna, Kreindler, & Olken, 2017)
Not only did congestion jump on those central Jakarta roads where carpooling
was previously mandated, it increased during times and in areas that were
never subject to the rule in the first place. In the hour following the evening
peak, for example (19:00-20:00), the repeal of 3-in-1 coincided with a roughly 50%
increase delays. The results for mid-day delays was were mixed, with an increase in
congestion between 0 and 30%. (Hanna, Kreindler, & Olken, 2017)
Finally, the repeal of Jakartas 3-in-1 policy resulted in increased congestion

not just on arterial roads; it also had a detrimental effect on secondary roads.
Two of the routes studied in detailed saw increases in delays of up to 27%, depend-
ing on the route and time of day. (Hanna, Kreindler, & Olken, 2017)
32 Unlocking Cities

The Jakarta story, while not ridesharing as defined in this paper, suggest both prac-
tical and arguably attainable results for ridesharing providers. While pooled riders
remain the minority in most rideshare markets, their share of total rides seems to
be growing. Lyft, for example, reported in 2015 that the pooled offering represented
50% of total Lyft trips in San Francisco and 30% of total Lyft trips in New York City.
(Lyft Blog, 2015) In Southeast Asia, Ubers pooled option represented approximate-
ly 25% of total trips in August 2017. (Uber Data, Aug 2017)
Benefit 3: Greater vehicle utilization per KM
In cities where taxis provides a substantial share of modality, ridesharing potential-

ly generates benefits from having fewer wasted kilometres compared with taxis.
Typically, taxis and rideshare vehicles spend only a fraction of their time on the
road actually conveying passengers. The remainder of the time a taxi or rideshare
vehicle is active is spent sitting in wait of a call or roaming the area looking for pas-
sengers. This non-productive travel is sometimes called dead kilometres.
All other things equal, a higher vehicle utilization i.e. fewer dead kilometres is a
good thing. It means that a fewer number of vehicles are needed to serve a commu-
nity, this reducing congestion. In this measurement of utilization, rideshare com-
pares quite favourably relative to taxis across multiple markets. Research by the
San Francisco County Transportation Authority (SFCTA) and Northeastern Univer-
sity indicated that for trips within San Francisco, rideshare vehicles demonstrate
approximately half of the dead kilometres (as a percentage of total KM) com-
pared to taxis. (San Francisco County Transportaton Authority, 2017). Similar re-
search by the US National Bureau of Economic Research (NBER) reached a similar
conclusion for San Francisco, Boston, Los Angeles, and Seattle that utilization of
Uber vehicles was approximately 40% greater than that of taxis. (Cramer & Krueger,
2016). Of the cities covered in the latter study, only in New York showed utilization
roughly equivalent between rideshare and taxis.
Benefit 4: Complementing public transport to accelerate adoption
One study of rideshare users across various U.S. cities found that after those sur-
veyed started using shared-use mobility,3 43% reported an increase in public transit
use, while 28% reported using public transit less. (Feigon & Murphy, 2016). The in-
crease in public transport usage may correlate with a car-light lifestyle, as the in-
crease in public transport and shared mobility usage tends to be higher for late-
night/weekend trips, when alcohol is involved, or in areas where public transit may
not be readily available. In these instances, ridesharing may serve as a means for
last-mile transport.
One way public transit systems seek to utilise ridesharing as a feeder mechanism is
by linking the booking or payment systems for both modes. (Feigon & Murphy,
2016) In Portland, Oregon, for example, a local transit authority (TriMet) has inte-
grated rideshare and car share booking capabilities into its public transit app. A
spokesperson for the department explained the decision: One of the things we're
trying to solve are the first and last mileThese are people we can't serve, finan-
cially. We wanted to provide other ride options that work really closely in synch
with transit. (Nijus, 2016)

Benefit 5: Helping optimise infrastructure timing & location
Another benefit provided by existing rideshare models is improved coverage of out-

lying areas. Studies of U.S. markets suggest rideshare provides greater coverage of
non-core city areas relative to taxis. Studies examining rideshare networks in Man-
hattan found that while the large majority of rideshare trips take place in those cit-
ies CBD areas, outlying areas were generally better served by rideshare, than
by taxis. (Schaller, 2017)
By operating in areas previously underserved by taxis and public transit, rideshare

provides the potential for reduced car use among drivers and greater access to pub-
lic transportation for those households without access to a vehicle. One U.S.-based
progressive advocacy group has argued that rideshare services in such areas should
be publicly subsidized so as to increase mobility literal and economic for low-in-
come households that would otherwise have limited access to existing public transit
corridors. (DeGood & Schwartz, 2016) Supporting this proposition, researchers sur-
veying various U.S. transit agencies reported that the agencies most interested in
complementary mobility options were those agencies with dispersed ridership, few-
er fixed guideway routes, or a higher proportion of relatively expensive operations
(such as paratransit), though the authors note increased contention with regard to
ridesharing (UberX, Lyft, etc.) specifically. (Feigon & Murphy, 2016)
While such a plan may seem far-fetched, some municipalities have indeed begun di-
verting funds from transportation-related projects and infrastructure to rideshare
companies. In 2016, a New Jersey suburb decided to subsidize Uber rides to its local
public transit hub, instead of using the funds to expand parking at the location.
(Fung, 2017) Numerous other transit departments have struck deals with rideshar-
ing companies, most often to provide bus-like services to otherwise underserved ar-
eas. (Brustein, 2016)
Despite the clear possible benefits of ridesharing, concerns have emerged about the
interaction between ridesharing and other transport modes such as taxi operators
and public transport players. BCG has therefore explored these concerns and poten-
tial ways forward. From our assessment, we found that a net positive outcome can
be realized for all stakeholders ridesharing is not and need not be a "zero sum"
game.To achieve net positive benefits to Asian cities, several conditions must be
achieved regarding:
Ridesharing substituting against private vehicles
Ridesharing benefits are obtained by providing greater transport efficiency (peo-

ple-kilometres) compared to private vehicles. However, to provide net positive ben-
efits for congestion, cities must ensure substitution of ridesharing for private vehi-
cles (private cars or motorcycles) and not public transport. While there is evidence
that rideshare can supplement public transport and support car-light lifestyles (see
above), there is mixed evidence suggestingthat ridesharing may substitute for pub-
lic transport use under certain conditions.4
34 Unlocking Cities

This challenge is potentially most significant for Tier I cities in Asia that currently
rely heavily on public transport. However, among the Tier I cities studied, the price
differential between private vehicle ownership and public transport is large given
government control over vehicle prices. Therefore, assuming rideshare prices re-
main more attractive in comparison to car ownership than public transport, the risk
of public transport substitution may not be significant.
This risk can be further mitigated by rideshare platforms and governments working
together to establish programs that make ridesharing services an appealing comple-
ment to public transport. For example, governments can work with ridesharing
platforms to provide commuters with live inter-modal travel data and to establish
discounts or pooling schemes for feeder transport to arterial public transport infra-
structure.
Utilisation of taxis
The rise of rideshare has been perceived to reduce taxi ridership in some cities. For
example, data from the Land Transport Authority of Singapore suggests that the
proportion of taxis sitting idle in yards has increased from 2016 to 2017.5 However,
the Ministry of Transport in Singapore has also suggested that rideshare has served
as a positive complement to taxis, particularly in peak hours.6, 7 In addition, the
emergence of rideshare technologies may have encouraged taxis to adopt more so-
phisticated technological advancements such as electronic applications, dynamic
pricing and smart supply-demand matching tools enhancing their competitive po-
sition and ultimately benefitting commuters. Furthermore, in Sydney, taxi ridership
has grown since the entrance of rideshare, suggesting that the risk of disruption to
taxis is uncertain and market specific.
Governments can also play a role in ensuring taxi companies improve their compet-
itive position while offering commuters better outcomes.For example,taxis should
be able to access the same technologies available to ridesharing vehicles. Both taxis
and private vehicles can form part of the flexible supply base necessary to realise
the congestion benefits outlined above. In particular, governments should ensure
that taxis can use apps to connect with passengers, and ensure that taxis can avail
themselves of supply-demand matching mechanisms such as dynamic pricing.
Partnerships between rideshare platforms and taxi companies can also benefit taxi
drivers. Recent examples of partnerships between rideshare platforms and taxi
companies include UberTAXI in Taiwan, UberFLASH in Malaysia and Grabs part-
nerships with multiple Singaporean and Vietnamese taxi companies. These part-
nerships promise to benefit taxi drivers by offering them access to technology
which may allow more responsive matching of supply to demand, thereby increas-
ing vehicle utilisation and ridership. These partnerships also benefit drivers by of-
fering access to large networks of potential passengers.

We believe net positive outcomes can be realized across stakeholders in the trans-
port landscape. Demand for transport will continue to grow across Asian cities,
leading to opportunities for incumbent transport models to evolve and for new
transport models to enter ultimately leading to better transport outcomes for
commuters.
2. Quantitative analysis
The focus of the quantitative analysis is to assess the impact of ridesharing on road
congestion under different scenarios of rideshare adoption. We define congestion as
the percentage of time difference in traveling during peak and non-peak hours com-
pared to the time it would take to travel the same distance at posted speed limits.
We have assessed peak hours at 7-9AM and 6-8pm.
Road congestion is driven by a set of elements:
1. Travel speed (actual and speed limit)
2. Road capacity in terms of number of total vehicles on the road
3. Traffic volume on road during the defined periods (peak, non-peak hours)
Appendix exhibit X: Road congestion driver tree

APPENDIX EXHIBIT X: ROAD CONGESTION DRIVER TREE
0
Road congestion
f
1 2
Actual drive-speed Post speed limit
1
f
3 4
Trac volume on road Road capacity
5 6
Passenger car
# vehicles by type equivalent conversion
7 8 9
Total people-KM demand Average occupancy by
per vehicle type Annual KM per vehicle
vehicle type
10 11
Total people-KM demand Modality share
of the city per transport mode
36 Unlocking Cities

Further details on each metric are below:
Appendix exhibit xi: Road EXHIBIT

APPENDIX congestion
XI: ROADdriver tree description
CONGESTION DRIVER TREE DESCRIPTION
Metrics Description Data Source
0
Road congestion % of additional travel time on average in peak, non-peak Tom Tom Trac Index
hours, when compared to driving at post speed limit Government statistics
Tom Tom trac data

1
Actual drive speed of vehicles on the road in peak, Google Map API
Actual drive speed
non-peak hours Government statistics
UBER travel data
2 Government data
Post speed limit Post speed limits on highways, urban roads per city
Press search
Academic studies on
3 Transportation
Total trac measured in passenger car equivalent units
Trac volume on road Engineering
on the road in peak, non-peak hours
Government statistics
UBER data
Estimated number of vehicles (in passenger car equivalent

4 Expert interviews on
units) that a single lane can throughput by type of road
Road capacity typical design throughput
- Highway: 2000 vehicles/link/lane
per lane by type of road
- Urban road: 1200 vehicles/link/lane
5 Number of vehicles by type: private cars, buses, taxi, Government statistics

# vehicles by type motorcycles, ridesharing cars and etc. UBER data
Academic studies on
6 Vehicle units used to convert dierent types of vehicles
Transportation
Passenger car equivalent to standard car unit based on the size/volume taken of
Engineering
a vehicle on the road
Expert interviews
7 Total distance travelled by the population using each of Government statistics

Total people-KM demand
per vehicle type the modes of transport Survey
8 Average total kilometers travelled annually per type of Government statistics

Annual KM per vehicle vehicle (private car, taxi, motorcycles and ridesharing car) Survey
9 Government statistics
Average occupancy by Average number of people in a vehicle per trip Survey
vehicle type Expert interviews
10 Total people-KM Total distance travelled by all modes of transport by

Government statistics
demand of the city total population of the city
11 Modality share per % of KMs travelled by each mode of transport Government statistics
transport mode

Key Findings for 2017 baseline
Road congestion in peak hours among the cities studied averages at 55%, with cer-
tain cities such as Bangkok, Manila, and Ho Chi Minh exceeding more than 100%.
This means that, on average, commuters take 55% longer to travel a given distance
in peak hours compared to if they travelled at posted speed limits.
AppendixAPPENDIX
exhibit XII: Current
EXHIBIT road
XII: CURRENT congestion
ROAD CONGESTION during peak
DURING PEAK hours
HOURS across
ACROSS cities
CITIES in 2017
IN 2017
150
100
Asia avg
132 134 67%
105 112
50
79
63 65 68 70
57
0
Singapore Hong Kong Surabaya Kuala Taipei Jakarta Bangkok Ho Chi Minh Manila Hanoi
Lumpur
Vehicle 0.2% 3.4% 6.3% 7.7% 0.8% 10.0% 6.4% 6.4% 4.0% 10.6%
growth (%)1
1. From 2011-2016 where data available from published government statistics 2. Peak hours dened as 7-9am, 6-8pm
1. From 2011-2016 where data available from published government statistics 2. Peak hours defined as 7-9am, 6-8pm
Note: Asia average taken from average of East Asian cities based on TomTom trac index Source: TomTom trac index;
Note: Asia average taken from average of East Asian cities based on TomTom traffic index Source: TomTom traffic index;
GoogleGoogle API;Government
API; Uber; Uber; Government statistics;
statistics; BCG analysis
BCG analysis
Indeed, we find that during peak hours, road capacity across all cities is in excess of
capacity to allow travel at posted speed limits.
Ho Chi
Kuala
Singapore Hong Kong Taipei Bangkok Surabaya Jakarta Minh Manila Hanoi
Lumpur
City
Vehicles in
excess of 40% 43% 47% 46% 62% 44% 51% 65% 72% 73%
capacity
38 Unlocking Cities

Public transport adoption, particularly rail, is key to managing road congestion in
cities. However, the share of transportation KM conveyed by public transport varies
greatly between cities.
Appendix exhibit XIII:

APPENDIX Current
EXHIBIT mileage
XIII: CURRENT modality
MILEAGE share
MODALITY byBYvehicle
SHARE typeINin
VEHICLE TYPE 2017
2017
% motorized modality share by est. passenger KMs travelled
100
Ridesharing
80
60 Motorbikes
40 81
58
51 Taxi
20 37 Private Cars
23
13 17 Public
4 9 10 transport
0
Hong Kong Singapore Taipei Kuala Bangkok Ho Chi Jakarta Manila Surabaya Hanoi
Lumpur Minh City
Modality
share, Mean 59% 18% 18%
public
transport Median 58% 18% 0%
Source: Government statistics; press search; commuter surveys; BCG Analysis

While public transportation adoption is highly linked to congestion, the efficiency

of vehicles used to provide transportation is also critical in assessing congestion. We
define efficiency based on the total people-kilometres each vehicle supplies per an-
num. This metric is driven by the total annual kilometres attributed to each trans-
port mode, the number of vehicles supporting each transport mode in the city, and
the average occupancy of each vehicle type, which corresponds to the ridership of
that transport mode.
The ability for ridesharing vehicles to provide greater transportation benefit de-
pends on the difference in people-kilometres each rideshare vehicle provides in
comparison to other modes of transport. To estimate people-kilometres, we used
available information on relative kilometres travelled per vehicle for ridesharing in
Singapore vs. taxis. In Singapore, based on available information, rideshare cars
travel 1/38 the kilometres of taxis per annum. We then extrapolated this across
Asian markets compared with taxi kilometres in each market. Despite the relatively
low kilometres travelled compared with taxis, ridesharing is still substantially more
efficient compared to private vehicles.

The figure below compares the estimated people-kilometres provided by rideshar-
ing vehicles against the #1 private vehicle transport in each city. The #1 private ve-
hicle mode varies between private cars and motorcycles across cities. Based on our
estimate, ridesharing is at least 1.3x more efficient than a privately owned vehicle,
and in highly congested cities such as Hanoi and Jakarta, ridesharing provides ~3x
greater people kilometres per vehicle.
Appendix exhibit XIV: Average annual people-kilometres travelled per vehicle type
APPENDIX EXHIBIT XIV: AVERAGE ANNUAL PEOPLE-KILOMETERS PER VEHICLE TYPE
Rideshare vehicles vs. #1 preferred mode of privately owned vehicle (car or motorbike)
RIDESHARE VEHICLES VS. # 1 PREFERRED MODE OF PRIVATELY OWNED VEHICLE (CAR OR MOTORBIKE)
People KM per vehicle, per annum
80,000
3.4x
1.7x 2.7x
60,000
3.2x
1.3x
1.8x 1.8x
40,000 1.9x 2.0x
1.7x
20,000
0
Hong Kong Singapore Manila Bangkok Taipei Kuala Lumpur Jakarta Ho Chi Surabaya Hanoi
Minh City
Private Motorbike Private Car Rideshare

Assessment of rideshare benefits in 2017

To assess the potential benefit of rideshare vehicles, we quantified the number of
vehicles that could be taken off the road in a scenario where the most widely pri-
vately owned vehicles were substituted by ridesharing. For example, in a market
where private cars provide the second highest form of modality and ridesharing
provides the fifth highest form of modality, we assessed how many vehicles could
be saved if ridesharing became the second highest form of modality.
40 Unlocking Cities

Under this scenario, ridesharing reduces the number of vehicles required at a sig-
nificant rate. In cities where private cars make up the majority of private transport,
between ~40%-60% of cars can be removed. In cities where motorcycles make up
the majority of private transport, between 55%-73% of motorcycles could be re-
placed by ridesharing.
Appendix exhibit XV: Percentage of private vehicles (car and motorcycle) and total vehi-
cles reduced with rideshare APPENDIX EXHIBIT XV:
PERCENTAGE OF PRIVATE VEHICLES (CAR AND MOTORCYCLE) AND TOTAL VEHICLES REDUCED WITH RIDESHARE
100%
80%
73% 71%
70%
66%
63% 63%
60%
60% 57% 56% 55%
53%
46% 46%
42%
39% 39%
40% 35%
31%
24%
20%
11%
0%
Singapore Hong Kong Taipei Kuala Bangkok Manila Jakarta Ho Chi Hanoi Surabaya
Lumpur Minh City
Total #
vehicles2
(million) 0.7 0.7 1.6 6 9.6 2.5 22 8 6 2.2
Total #
private 0.5 0.6 0.6 4 6 1.5 4 7.8 5.6 1.8
vehicles 3
(million)
Car Motorcycles
% of private vehicles % of total vehicles % of private motorcycle
1. With rideshare scenario under which ridesharing replaces private cars as the #2 or #3 mode of transport in respective cities and pool
1. Withconstitutes
rideshare50%
scenario under
of rides which
2. Total ridesharing
number replaces
of vehicles private
includes cars
private asmotorcycles,
cars, the #2 or #3buses,
modetaxi
of and
transport in respective
rideshare cities
cars, 3. Total and pool
number of cars
include
constitutes private
50% cars 2.
of rides and ridesharing
Total numbercars.
of vehicles includes private cars, motorcycles, buses, taxi and rideshare cars, 3. Total number of cars
include private
Source: cars and ridesharing
Government cars.
statistics; BCG Analysis
Source: Government statistics; BCG Analysis

28
As a result, in this scenario, congestion is also estimated to decline as a result in this
reduction of vehicles due to rideshare adoption.
Appendix exhibit XVI: Road congestion during peak hours, before and after rideshare im-
pact under scenario
APPENDIX where
EXHIBIT rideshare
XVI: ROAD substitutes
CONGESTION for
DURING PEAK #1 mode
HOURS, ofAND
BEFORE private
AFTER vehicle
RIDESHARE
IMPACT UNDER SCENARIO WHERE RIDESHARE SUBSTITUTES FOR #1 MODE OF PRIVATE VEHICLE
Peak congestion %
150
-88% -85%
-85%
-92%
100
-51%
-91% -81%
-90% -72%
-77%
50
0
Singapore Hong Kong Kuala Taipei Jakarta Bangkok Manila Surabaya Ho Chi Hanoi
Lumpur Minh City
No. of private
vehicles reduced 0.2 0.3 2.5 0.4 2.5 3.7 1 0.7 2.4 3.5
after rideshare
(million)
Car Motorcycles
Before rideshare After rideshare
Note: Reductions in congestion based on high-adoption, high-pooling scenario

Reductions
Note: Source: in congestion based on high-adoption, high-pooling scenario
BCG analysis
Assessing the benefits of rideshare for 2022

To assess the impact of rideshare in 2022, we first had to estimate the increase in
transportation demand between 2017 and 2022. Research by the National Center
for Sustainable Transportation (Circella, Tiedeman, Handy, & Mokhtarian, 2015)
shows a strong correlation between transportation demand and wealth. It suggests
that on a per capita basis, people tend to travel more as wealth increases.
42 Unlocking Cities

These results are reinforced by BCG research into economic growth and its effect on
passenger land transportation in OECD member nations, between 1970 and 2015.
Our analysis suggests an approximate 1:0.75 relationship between annual GDP
growth per capita and annual growth in land transport passenger kilometres. We
therefore project demand for transport to grow by around 20% across cities in five
years. This growth is driven by increased city population as well as wealth.
Appendix exhibit XVII: AnnualAPPENDIX

travel demand ofANNUAL
EXHIBIT XVII: the city in 2017 and 2022
TRAVEL
DEMAND OF THE CITY IN 2017 AND 2022
Annual travel demand in billion kilometers
250
200
150
100
50
0
Taipei Singapore Hong Kong Kuala Bangkok Surabaya Manila Hanoi Jakarta Ho Chi
Lumpur Minh City
2017 total KM 2022 total KM

Note: 2022 total KM is forecasted based on population and wealth growth.
2022 total
Note:Source: KMCenter
National is forecasted based on
for Sustainable population and
Transportation; wealthIntelligence
Economist growth. Unit; BCG analysis
Source: National Center for Sustainable Transportation; Economist Intelligence Unit; BCG analysis
Based on this projected demand and historical vehicle growth in each city, we esti-
mate that congestion during peak hours will worsen in a number of cities, assuming
that modality and vehicle utilization remain the same as 2017. For Singapore, we
have assumed car growth will be controlled by government mechanisms to main-
tain congestion. For Taipei, the congestion will likely decrease due to negative his-
torical vehicle growth.

30
Appendix APPENDIX
exhibitEXHIBIT
XVIII: XVIII: ESTIMATED
Estimated ROADcongestion
road CONGESTION DURING
duringPEAK
peakHOURS IN 2022
hours VS. 2017
in 2022 vs. 2017
P eak congestion (%)
400 +177%
+204% +152%
300
+93%
200 +108%
371
340 333
+94%
+47%
+46%
100 +4% -16% 203
164
126 132 134
100 105 112
92 79
57 59 63 70 59 68 65
0
Singapore Hong Kong Taipei Kuala Surabaya Bangkok Jakarta Ho Chi Manila Hanoi
Lumpur Minh City
2017 peak congestion level 2022 peak congestion level
Note: 2022 congestion is forecasted based on trac volume increase, which is in line with travel demand increase, Assumptions include same
Note: 2022 congestion is forecasted based on traffic volume increase, which is in line with travel demand increase, Assumptions include
modality share and vehicle utilization as 2017.
sameSource:
modality
Tomshare and vehicle
Tom Trac utilization
Index; Google Map as
API;2017.
Economist Intelligence Unit; BCG analysis
Source: Tom Tom Traffic Index; Google Map API; Economist Intelligence Unit; BCG analysis
The increase in travel demand in the city could potentially be met by a greater
adoption of public transportation. However, we estimate that the required increase
in rail network infrastructure may be greater than the capacity which will come
online by 2022 in Tier 2 and Tier 3 cities.
44 Unlocking Cities

Appendix exhibit XIX: Estimated % people-kilometres that must be travelled by public
transportation toEXHIBIT
APPENDIX maintain congestion
XIX: ROAD levels
CONGESTION vs.PEAK
DURING estimated public
HOURS BEFORE transportation
VS. AFTER capacity
RIDESHARE (2017)
% KM travelled by public transport

100
80
60
40
20
0
Singapore Hong Kong Taipei Kuala Bangkok Ho Chi Manila Jakarta Surabaya
Lumpur Minh City
2017 % KM travelled by public transport

2022 % KM travelled by public transport to maintain current peak congestion
Estimated public transport capacity by 20221
Investment
US$ Billion 20 25 14 6 24 10 51 5 0.09
(2017 2022)
1.Capacity is estimated based on current rail network and new rail lines/existing line extensions in operation before 2022 in each city: Thomson East
1.Capacity is estimated
Coast Line, Downtown based
line 3on current for
extension railSingapore;
network andtotalnew
new rail lines/existing
railway line extensions
projects equivalent to 25% ofin operation
current before
capacity 2022Kong;
in Hong in each city: Line
Circular
Thomson
stageEast CoastLine,
1, Anking Line, Downtown
Danhai line 3 Line
LRT, Wanda extension
stage 1,for Singapore;
Xinzhuang Linetotal new railway
extension projects
for Taipei; equivalent
MRT Line to 25%
2 for Kuala of current
Lumpur; 10 newcapacity
rail lines in
and 3
Hongexisting
Kong; Circular Line stage
line extensions 1, Anking
for Bangkok; rstLine,
MetroDanhai
Line andLRT, Wanda
3 LRT linesLine stage 1,
for Jakarta; Xinzhuang
6 Metro Line 109
Rails (total extension
KM) forfor
HoTaipei;
Chi MinhMRT Linerailway
; 6 new 2 for Kuala
lines
Lumpur;
(total10246
new railforlines
KM) andone
Manila; 3 existing
monorailline
for extensions
Surabaya for Bangkok; first Metro Line and 3 LRT lines for Jakarta; 6 Metro Rails (total 109
KM) for Ho Chi
Source: Minh ; 6 announcement
Government new railway lines (total 246infrastructure
on transport KM) for Manila;masterone monorail
plan; for Surabaya
BCG analysis
Source: Government announcement on transport infrastructure master plan; BCG analysis
We therefore estimate that rideshare could play a role complementing public

transportation. In tier 1 cities, rideshare can help alleviate pressure on public
transport, which may be strained by growing demand. In Tier 2 and 3 cities,
rideshare can complement greater adoption of public transport to maintain
congestion. We estimate that the growth in rideshare adoption could support cities
by lowering or maintaining congestion.

Appendix exhibit XX: Increase
APPENDIXin rideshare
EXHIBIT adoption
XX: INCREASE needed
IN RIDESHARE to maintain or reduce con-
ADOPTION
gestion NEEDED TO MAINTAIN OR REDUCE CONGESTION
2022 Congestion 2022 Congestion

without rideshare1 with rideshare 2
Singapore 59% 34%
Hong Kong 92% 36%
Taipei 59% 40%
Kuala Lumpur 100% 68%
Bangkok 203% 105%
Jakarta 164% 79%
Ho Chi 340% 112%

Minh City
Manila 333% 132%
126% 65%
Surabaya
0 10 20 30 50 % modal share by rideshare

Modal share of rideshare in 2017 Modal share of rideshare in 2022
1. Forecasted based on trac volume increase, which is in line with travel demand increase, Assumptions include same modality share and vehicle
1. Forecasted based
utilization as 2017.on traffic volume increase, which is in line with travel demand increase, Assumptions include same modality share and
vehicle utilization
2. Assuming as 2017.
rideshare complements public transport in tier 1 cities to reach all-day average congestion level, and in tier 2 and 3 cities to maintain
2. Assuming
current peakrideshare complements public transport in tier 1 cities to reach all-day average congestion level, and in tier 2 and 3 cities to
congestion.
maintain
Source:current peak congestion.
BCG analysis
Substantial increases in adoption of rideshare vehicles are therefore needed to

achieve the benefits associated with substitution against private vehicle ownership.
Supply caps, which can occur in the form of outright caps on private car hire
vehicles and restrictions on driver recruitment, can therefore be a barrier to
achieving these benefits. For example, in Hong Kong, stated supply caps of 1,500
private cars9 is approximately 70x lower than the estimated number needed to
reduce congestion by half while maintaining current levels of public transport
adoption. In Ho Chi Minh City where the government has announced plans to limit
contract cars to 12,700 by 2020, the cap (assuming the limit is taken entirely by
contract cars) would need to rise an enormous 170x to accommodate the number
of rideshare vehicles needed to achieve the benefits articulated above.
46 Unlocking Cities

Assessing the broader benefits of rideshare
Cities that rely heavily on private cars also require substantial parking infrastruc-
ture to support this mode of transport. We estimate that the space needed to ac-
commodate cars in each city studied ranges from 1 million to 25 million hectares.
This space is substantial for example, in Jakarta, the estimated space needed for
parking is equivalent to 24 thousand football fields.
Appendix exhibit XXI: Estimated

APPENDIX space needed
EXHIBIT XXI: ESTIMATED to serve
SPACE NEEDED private
TO SERVE carsCARS
PRIVATE in 2017
IN 2017
Hectares (k)
20
24.8
10
16.8 17.6
6.0
2.1 2.4 2.6 1.5 1.3 1.7
0
Singapore Hong Kong Taipei Kuala Bangkok Jakarta Hanoi Ho Chi Manila Surabaya
Lumpur Minh City
Lots(k) 1,100 1,253 1,341 8,829 12,995 9,247 802 705 3,165 896
Note: Area represents estimated total flat area of all parking lots (existing and needed) to serve a citys car population. Area estimated by
Note: Area represents estimated total at area of all parking lots (existing and needed) to serve a city's car population. Area estimated by deriving
derivingratio
ratioofofcars
cars(private
(private + rideshare)
+ rideshare) to estimated
to estimated parking
parking lots
lots in in Singapore
Singapore (~2.2) (~2.2) andextrapolating
and then then extrapolating
this ratiothis ratiopopulations
to car to car populations
in other
in othermarkets.
markets.Assumes
Assumes standard
standard parking
parking lot (19m2).
lot (19m2).
HDB, LTA,
Source:Source: HDB,URA,
LTA, expert interviews,
URA, expert BCGBCG
interviews, analysis
analysis

Higher rideshare adoption could result in a reduction in cars. As a result, parking
infrastructure currently used for parking could also be re-purposed. We estimate
that up to 16 thousand hectares of space could be re-purposed driven by the
reduction in cars.
Appendix exhibit XXII: Estimated space that can be saved by adopting rideshare assuming
APPENDIX
rideshare substitutes forEXHIBIT
privateXXII: ESTIMATED SPACE THAT CAN BE SAVED BY ADOPTING
cars
RIDESHARE ASSUMING RIDESHARE SUBSTITUTES FOR PRIVATE CARS
New slide (same as exhibit IX)
Hectares saved with rideshare Landmark equivalent
Singapore 872 2x Sentosa
Hong Kong 1,264 67x Victoria Park
Taipei 1,619 197x Botanic Gardens
Kuala Lumpur 9,583 139x Lake Gardens
Bangkok 15,556 273x Lumpini Park
Jakarta 10,647 6x Soekarno -Hatta Airport
Hanoi 339 4x Old Quarter
Ho Chi 366 17x Zoo 1

Minh City
Manila 3,362 26x EDSA
Surabaya 545 1x Tanjung Priok
0 10,000 15,000 20,000

Hectares
1. HCMC
1. HCMCZooZoo
andandGarden complex
Garden complex
Note: SizeSize
Note: of local landmarks
of local landmarksvary
vary greatly betweencities.
greatly between cities. Area
Area represents
represents estimated
estimated total total flatofarea
at area of all parking
all parking lots and
lots (existing (existing
needed)andto needed)
serve
to serve a citys
a city's car population.
car population. AreaArea estimated
estimated by deriving
by deriving ratio ofratio
carsof cars (private
(private + rideshare)
+ rideshare) to estimated
to estimated parking
parking lots lots in Singapore
in Singapore (~2.2) and (~2.2)
then
extrapolating
and then this ratio
extrapolating to car
this populations
ratio in other markets.
to car populations in otherAssumes
markets. standard parking
Assumes lot (19m2),
standard Area lot
parking saved underArea
(19m2), hypothetical scenario
saved under in which
hypothetical
rideshare
scenario becomes
in which displaces
rideshare private displaces
becomes vehicles in private
terms ofvehicles
modal split and 50%
in terms ofofmodal
rideshare
splitisand
pooling.
50% of rideshare is pooling.
Source:
Source: ASEANASEAN Maritime
Maritime WorkingGroup,
Working Group,Data.Gov.Sg,
Data.Gov.Sg, FIFA,
FIFA, MapDevelopers/Google
MapDevelopers/Google Maps, HDB,
Maps, HK HK
HDB, Census and Statistics
Census Dept.,Dept.,
and Statistics LTA, Manila
LTA,
Times,
Manila Perdana
Times, Botanical
Perdana Garden,
Botanical URA, Thanhnien
Garden, News, The
URA, Thanhnien Straits
News, TheTimes, Taipei
Straits Botanical
Times, TaipeiGarden, expert
Botanical interviews,
Garden, BCGinterviews,
expert analysis BCG
analysis
48 Unlocking Cities

Notes
1. Authors of the study note that the non-probability (i.e. opt-in) methodology employed provides an
insightful example of consumer behaviour in response to rideshare suspension, but not one that
should be extrapolated to all Austin rideshare users
2. During this period, other rideshare platforms continued to operate within Austin. The portion of
former Uber and Lyft customers who migrated to this platforms is roughly the same as the portion
that migrated to cars (roughly 40%). One can logically conclude that if these alternate rideshare
options were not available, the shift to private vehicles and increase in car ownership would have been
greater than those observed in the study
3. The paper referenced used shared-use modes to include public transit, rideshare (e.g. Uber, Lyft,
etc.), bike share, and car sharing (e.g. Zipcar)
4. For example, in the US, according to the National Academy of Sciences study of shared mobility
users referenced earlier, 43% of individuals reported an increase in their use of public transport, while
only 28% of individuals reported using public transport less. However, a study from UC Davis
Disruptive Transportation: The Adoption, Utilization, and Impacts of Ride-Hailing in the United States
(October 2017)suggest that ridesharing may decrease use of bus and light-rail services by 6% and 3%
respectively in several major metropolitan US cities
5. The proportion of Singapore taxis sitting idle in yards increased roughly 80% in the first five months
of 2017 over the same period in 2016
6. Straits Times article indicates taxi trips have declined from 967K daily taxi trips from 2013 to 954K
in 2016 - Straits Times February 2017
7. Channel News Asia Article titled: Dont stop taxi industry from adapting to competition: Ng Chee
Meng - Channel News Asia April 2017
8. http://www.straitstimes.com/singapore/transport/private-hire-cars-may-be-adding-to-congestion
9. Road Traffic (Public Service Vehicles) Regulations,regulation 19(1)) for private hire car service

50 Unlocking Cities
To find the latest BCG content and register to receive e-alerts on this topic or others, please visit bcgperspectives.com.
Follow bcg.perspectives on Facebook and Twitter.
The Boston Consulting Group, Inc. 2017. All rights reserved.

11/17
Abu Dhabi Chicago Kiev Munich Shanghai
Amsterdam Cologne Kuala Lumpur Nagoya Singapore
Athens Copenhagen Lagos New Delhi Stockholm
Atlanta Dallas Lima New Jersey Stuttgart
Auckland Denver Lisbon New York Sydney
Bangkok Detroit London Oslo Taipei
Barcelona Dubai Los Angeles Paris Tel Aviv
Beijing Dsseldorf Luanda Perth Tokyo
Berlin Frankfurt Madrid Philadelphia Toronto
Bogot Geneva Melbourne Prague Vienna
Boston Hamburg Mexico City Rio de Janeiro Warsaw
Brussels Helsinki Miami Riyadh Washington
Budapest Ho Chi Minh City Milan Rome Zurich
Buenos Aires Hong Kong Minneapolis San Francisco
Calgary Houston Monterrey Santiago
Canberra Istanbul Montral So Paulo
Casablanca Jakarta Moscow Seattle
Chennai Johannesburg Mumbai Seoul bcg.com

Unlocking Cities Report BCG Uber

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

Unlocking Cities Report BCG Uber

Uploaded by

Copyright:

Available Formats

Unlocking Cities

The impact of ridesharing in Southeast Asia and beyond

With these combined compo- Collaboration between

Increase occupancy per

The Boston Consulting Group 3

Growth in population and wealth have led to an explosion in transport demand in

Supporting car-light aspirations and lifestyles: 10-40% of commuters who

Increasing occupancy per vehicle: Pooling of commuters can raise occupancy

Complementing public transport adoption: Rideshare can support public

Optimising timing of infrastructure investment: Rideshare has been shown

Ridesharing substitution of private vehicles: Ridesharing benefits come

Utilisation of taxis: Ridesharing provides new possible avenues of income

Ridesharing must achieve critical mass to realise significant benefits. Rideshare

The Boston Consulting Group 5

Collaboration between rideshare and public sector: Encouraging rideshare

Ridesharing has the potential to positively contribute to the transport challenge

EXHIBIT I: INDEXED ESTIMATED GROWTH IN TRAVEL DEMAND (1980 = 100)

In response to this growing transport demand, governments in Asia have invested

Tier 1: Cities with well-developed and formally organised public transport,

Tier 2: Cities that have more recently developed extensive, formally

The Boston Consulting Group 7

Road and rail infrastructure rating1

% of KM travelled in city by public transport

Ridesharing and its implications for Asian cities

We define ridesharing as the combination of several elements:

Dynamic routing and smart matching technology: To efficiently map supply

The Boston Consulting Group 9

In Tier I cities, greater ridesharing adoption could benefit cities by:

Supporting car-light ambitions: In Singapore, Hong Kong and Taipei between

Alternatives to car ownership: Car ownership growth in both Kuala Lumpur

Supporting public transport adoption: Despite the significant investment in

Supplementing incomes: ~25% of drivers in Kuala Lumpur and Bangkok show

10 Unlocking Cities

Alternatives to car ownership: At current rates of vehicle growth, congestion

The Boston Consulting Group 11

To give a sense of the magnitude of the potential impact, we assessed a hypotheti-

12 Unlocking Cities

Private Motorbike Private Car Rideshare

% of cars % of total vehicles % of private motorcycle

The Boston Consulting Group 13

After rideshare Before rideshare

Exhibit IX: Estimated

Singapore 872 2x Sentosa

Hong Kong 1,264 67x Victoria Park

Taipei 1,619 197x Botanic Gardens

Kuala Lumpur 9,583 139x Lake Gardens

Bangkok 15,556 273x Lumpini Park

Jakarta 10,647 6x Soekarno-Hatta Airport

Hanoi 339 4x Old Quarter

Ho Chi Minh City 366 17x Zoo 1

Manila 3,362 26x EDSA

Surabaya 545 1x Tanjung Priok

0 10,000 15,000 20,000

14 Unlocking Cities

Tier I cities are planning significant expansion of public transport. Singapore,

The Boston Consulting Group 15

These benefits previously discussed are supported by recent studies conducted on

Benefit 1: Supporting a car-light lifestyle

16 Unlocking Cities

Benefit 3: Greater vehicle utilisation per kilometre

Benefit 4: Complementing public transport to accelerate adop-

The Boston Consulting Group 17

Ridesharing substituting against private vehicles

The Boston Consulting Group 3

The Boston Consulting Group 5

The Boston Consulting Group 7

The Boston Consulting Group 9

10 Unlocking Cities

The Boston Consulting Group 11

12 Unlocking Cities

The Boston Consulting Group 13

14 Unlocking Cities

The Boston Consulting Group 15

16 Unlocking Cities

The Boston Consulting Group 17

18 Unlocking Cities

The Boston Consulting Group 19

20 Unlocking Cities

The Boston Consulting Group 21

22 Unlocking Cities

The Boston Consulting Group 23

24 Unlocking Cities

The Boston Consulting Group 25

26 Unlocking Cities

The Boston Consulting Group 27

28 Unlocking Cities

The Boston Consulting Group 29

30 Unlocking Cities

The Boston Consulting Group 31

32 Unlocking Cities

The Boston Consulting Group 33

34 Unlocking Cities

The Boston Consulting Group 35

36 Unlocking Cities