This document summarizes research on the environmental impacts of shared mobility services in North America. It finds that carsharing reduces personal vehicle ownership, with estimates of 9-13 vehicles removed for each carsharing vehicle. Carsharing users drive less and bike and transit use increases. The growth of bikesharing systems worldwide and in the US is also summarized, with over 1 million public bikeshares now worldwide. Surveys find bikesharing users tend to be wealthier, more educated, younger and drive less as a result.

1. ENVIRONMENTAL
IMPACTS
OF
SHARED
MOBILITY:
INSIGHTS
FROM
NORTH
AMERICA
Susan
A.
Shaheen,
Ph.D.
Adjunct
Professor,
Civil
&
Environmental
Engineering
and
Co-­‐Director,
TSRC
University
of
California,
Berkeley

2. OVERVIEW
Ø
Carsharing
Ø
Market
trends
Ø
N.
American
survey
findings
Ø
Bikesharing
Ø Market
trends
Ø
N.
American
survey
findings
Ø
Ridesourcing/TNCs
Ø Market
trends
Ø San
Francisco
survey
findings
Ø
Acknowledgements

3. © UC Berkeley, 2015
Roundtrip
Carsharing:
Round
trip,
pay
by
the
hour/mile,
non-­‐
profit
and
for
profit
fleet
models
Peer-­‐to-­‐Peer
Carsharing:
Shared
use
of
private
vehicle
typically
managed
by
third
party
One-­‐Way
Carsharing:
Pay
by
the
minute,
point
to
point,
fleet
operated,
street
parking
agreements
Fractional
Ownership
Carsharing:
Individuals
sublease
or
subscribe
to
a
vehicle
owned
by
a
third
party
There
are
many
flavors
of
carsharing
Carsharing

4. GROWTH
OF
WORLDWIDE
CARSHARING
2006 2008 2010 2012 2014
Members 346,610 670,822 1,163,405 1,788,027 4,828,616
Vehicles 11,501 19,403 31,967 43,554 103,980
0
20,000
40,000
60,000
80,000
100,000
120,000
0
1,000,000
2,000,000
3,000,000
4,000,000
5,000,000
6,000,000
Vehicles
Members
© UC Berkeley, 2015
Shaheen
and
Cohen,
2015

5. NORTH
AMERICAN
LONGITUDINAL
TRENDS
2006 2008 2010 2012 2014
Members 117,656 318,898 516,100 908,584 1,625,652
Vehicles 3,337 7,505 10,420 15,795 24,210
0
5,000
10,000
15,000
20,000
25,000
30,000
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
Vehicles
Members
Shaheen
and
Cohen,
2015
© UC Berkeley,
2015

6. © UC Berkeley, 2015
Public
Bikesharing:
Point
to
point,
pay
by
the
½
hr,
fleet
operated,
docking
stations
Closed
Community
Bikesharing:
Campuses
and
closed
membership,
mainly
roundtrip,
linking
to
carsharing
Peer-­‐to-­‐Peer
Bikesharing:
Rent
or
borrow
hourly
or
daily
from
individuals
or
bike
rental
shops
Growing
exponentially
in
urban
centers
Bikesharing

7. WORLDWIDE
&
US
BIKESHARING:
May
2015
Ø 880
cities
with
IT-­‐based
operating
systems
Ø 1,036,000
bikes
Ø ~811,500
bikes
in
China
(and
256
cities)
Ø US:
72
cities
with
IT-­‐based
systems
Ø ~24,700
bikes
Ø 2,440
stations
Ø In
2015,
21
new
programs
to
begin
operating
in
world:
13
are
in
China
and
8
in
U.S.
Source:
Russell
Meddin,
2015

8.
2008
N.
AMERICAN
CARSHARING
SURVEY
Ø
Survey
implemented
from
Sept.
to
Nov.
2008
Ø
~9,500
completed
surveys;
analysis
based
on
6,281
hhds
Ø
Completion
rate
~80%
Ø
Online
survey
challenging
Ø Took
between
10
to
15
minutes
for
most
respondents
to
complete
Martin,
Shaheen,
Lidicker,
2010

9.
PARTICIPATING
ORGANIZATIONS
American
-­‐ City
CarShare
-­‐ CityWheels
-­‐ Community
Car
-­‐ Community
Carshare
of
Bellingham
-­‐ Igo
-­‐ PhillyCarShare
-­‐ Zipcar
Canadian
-­‐ AutoShare
-­‐ Communauto
-­‐ Co-­‐operative
Auto
Network
-­‐ VrtuCar
-­‐ Zipcar

10.
N.A.
VEHICLE
HOLDINGS:
KEY
FINDINGS
Ø
Between
9
to
13
vehicles
removed,
including
postponed
purchase
Ø
4
to
6
vehicles/carsharing
vehicle
sold
due
to
carsharing
Ø
Most
shift
due
to
1
car
households
becoming
carless
Ø
Second
largest
shift,
2
car
households
become
1
car
households
Ø
25%
sell
a
vehicle;
25%
postpone
purchase
Ø
Net
CO2
reduction
of
27%
observed
and
43%
full
impact
(average)
Martin,
Shaheen,
Lidicker,
2010

11. 2013
N.
AMERICAN
BIKESHARING
STUDY
Ø
23
operator
interviews
in
the
US,
Canada,
&
Mexico
Ø
IT-­‐based
bikesharing
program
growth
tracking
(including
planned
programs)
Ø
Surveys
with
members
of
5
public
bikesharing
systems
in
Spring/Summer
2013
Ø
2013
operational
data
analysis
from
Nice
Ride
MN
© UC Berkeley, 2015
Shaheen
et
al.,
2014

12. 2013
MEMBER
SURVEY:
DEMOGRAPHICS
Compared
to
general
population
bikesharing
users
tend
to
be…
Ø
Wealthier
Ø
More
educated
Ø
Younger
Ø
Caucasian
Ø
Male
Shaheen
et
al.,
2014
© UC Berkeley, 2015

13. MEMBER
SURVEY:
2013
Shaheen
et
al.,
2014
©
UC
Berkeley,
2015

14. CHANGE
IN
DRIVING
0% 0%
44%
9%
29%
4% 0%0% 0%
49%
6%
38%
0% 0%
0%
10%
20%
30%
40%
50%
60%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
drive
before/after
Changed
driving,
not
due
to
bikesharing.
As
a
result
of
my
use
of
bikesharing,
I
drive
a
personal
vehicle
(e.g.,
car,
SUV,
etc.)
…
Minneapolis
Saint-­‐Paul,
N
=
620
Salt
Lake
City,
N
=
72
0% 0%
19%
10%
16%
3%
0%0% 0%
23%
12%
24%
4%
0%
0%
10%
20%
30%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
drive
before/after
Changed
driving,
not
due
to
bikesharing.
Montreal,
N
=
1095
Toronto,
N
=
993
1% 0%
32%
21% 19%
2% 0%
0%
5%
10%
15%
20%
25%
30%
35%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
drive
before/after
Changed
driving,
not
due
to
bikesharing.
Mexico
City,
N
=
3329
Shaheen
et
al.,
2014
©
UC
Berkeley,
2015

15. 10%
15%
27%
29%
13%
1%
5%
14%
18%
25% 24%
13%
0%
7%
0%
5%
10%
15%
20%
25%
30%
35%
Less
than
once
a
month
Once
a
month Every
other
week
1
to
3
days
per
week
4
to
6
days
per
week
Once
a
day More
than
once
a
day
Currently,
how
often
do
you
check
out
a
bikesharing
bicycle?
Minneapolis
Saint-­‐Paul,
N
=
618
Salt
Lake
City,
N
=
72
5% 3%
10%
30%
24%
7%
22%20%
7%
13%
23%
19%
4%
15%
0%
10%
20%
30%
40%
Less
than
once
a
month
Once
a
month Every
other
week
1
to
3
days
per
week
4
to
6
days
per
week
Once
a
day More
than
once
a
day
Montreal,
N
=
1090
Toronto,
N
=
1010
4% 6%
12%
30%
21%
7%
21%
0%
10%
20%
30%
40%
Less
than
once
a
month
Once
a
month Every
other
week
1
to
3
days
per
week
4
to
6
days
per
week
Once
a
day More
than
once
a
day
Mexico
City,
N
=
3330
Shaheen
et
al.,
2014
©
UC
Berkeley,
2015

16. CHANGE
IN
URBAN
RAIL
Shaheen
et
al.,
2012
1%
10%
2% 0%
61%
3% 0%1%
13%
4% 3%
72%
0% 0%
0%
10%
20%
30%
40%
50%
60%
70%
80%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
rail
before/after
Changed
rail
use,
not
due
to
bikesharing.
As
a
result
of
my
use
of
bikesharing,
I
use
urban
rail…
Minneapolis
Saint-­‐Paul,
N
=
620
Salt
Lake
City,
N
=
72
1%
6%
37%
20%
30%
1% 0%1%
7%
35%
14%
39%
2% 0%
0%
10%
20%
30%
40%
50%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
rail
before/after
Changed
rail
use,
not
due
to
bikesharing.
Montreal,
N
=
1096
Toronto,
N
=
1005
4%
9% 12%
5%
44%
1% 0%
0%
10%
20%
30%
40%
50%
Much
more
often More
often Less
often Much
less
often About
the
same Did
not
ride
the
rail
before/after
Changed
rail
use,
not
due
to
bikesharing.
Mexico
City,
N
=
3333
Shaheen
et
al.,
2014
© UC Berkeley, 2015

17. Response
Categories Montreal Toronto Minneapolis-­‐Saint
Paul Salt
Lake
City Mexico
City
Lower
cost
and
faster
travel 25% 48% 0% 0% 28%
Just
lower
cost 5% 9% 7% 0% 2%
Too
many
connections
(not
have
to
transfer)
3% 2% 7% 0% 6%
Just
faster
travel 14% 14% 14% 40% 12%
Improve
travel
time
reliability
4% 7% 0% 60% 6%
Want
to
get
exercise 31% 8% 50% 0% 17%
Public
transit
vehicle
is
crowded
6% 6% 0% 0% 18%
No
space
for
my
bike,
which
I
use
to
connect
0% 0% 0% 0% 2%
I
consider
it
safer
to
travel
with
bikesharing
1% 0% 7% 0% 2%
Not
applicable 1% 2% 0% 0% 3%
Other,
please
specify: 8% 5% 14% 0% 3%
Total
N 631 491 14 5 577
What
is
the
primary
reason
that
you
are
using
the
rail
LESS
because
of
bikesharing?
Shaheen
et
al.,
2014
© UC Berkeley, 2015

18. © UC Berkeley, 2015
A
new
category
of
transportation
services;
need
for
study
Ridesourcing
Ridesourcing/TNC:
Prearranged
trips,
App
to
pay
and
connect
passengers
with
drivers
who
use
their
personal
vehicles

19. SOME
RIDESOURCING/E-­‐HAIL
MARKET
TRENDS
Ø Lyft:
60
cities;
51,000
drivers
(2014)
Ø Uber:
58
countries;
311
cities;
over
162,000
drivers
in
U.S.
Ø Sidecar:
10
cities;
~10,000
drivers
Ø Flywheel:
6
cities,
over
5,000
drivers
Ø Curb:
60
cities;
35,000
cabs
Said,
2015;
Miller,
2015
© UC Berkeley, 2015

20. RIDESOURCING:
SOME
EARLY
UNDERSTANDING
Taxi,
39%
Bus,
24%
Rail,
9%
Walk,
8%
Drive
Own
Car,
6%
Bike,
2%
Get
A
Ride,
1%
Other,
11%
Ø 92%
would
have
still
made
this
trip
Ø 8%
induced
travel
effect
Ø 33%
would
have
taken
public
transit
(bus
or
rail)
Ø 4%
named
a
transit
station
origin/destination,
suggesting
some
ridesourcing
usage
to
access
public
transportation
Ø 20%
avoided
driving
after
drinkings
Rayle
et
al.
2014
How
would
you
have
made
this
trip
if
Uber/Lyft/Sidecar
were
not
available?
© UC Berkeley, 2015

21. KEY
FINDINGS:
WAIT
TIMES
Rayle
et
al,
2014
About
how
long
did
you
wait
for
your
ride
(from
the
time
you
made
the
request
to
the
time
the
vehicle
arrived)?
Percentages
of
wait
times
less
than
or
equal
to
10
minutes:
Wait
Times
Ridesourcing
Taxi
(Phone)
Taxi
(Street
Hail)
M-­‐F
4am-­‐6pm
93%
35%
39%
M-­‐F
(6pm-­‐4am)
92%
16%
33%
S-­‐Su
88%
16%
25%
© UC Berkeley, 2015

22. KEY
FINDINGS:
VEHICLE
OWNERSHIP,
OCCUPANCY
&
DRIVING
FREQUENCY
Ø Ridesourcing
and
taxis
serve
residents
who
don’t
own
a
car
Ø Ridesourcing
survey:
43%
no
vehicle
at
home
Ø Taxi
survey:
35%
car-­‐less
Ø
Occupancy:
1.8
TNCs
and
1.1
taxis
Ø
Ridesourcing
is
still
new,
with
potential
to
impact
VMT/
VKT
and
vehicle
ownership
Ø
90%
of
vehicle
owners
did
not
change
ownership
level
Ø
40%
drove
less
since
using
ridesourcing
Rayle
et
al.
2014

23. ACKNOWLEDGEMENTS
Ø Mineta
Transportation
Institute,
San
Jose
State
University
Ø California
Department
of
Transportation
Ø Adam
Cohen,
Elliot
Martin,
Nelson
Chan,
and
Matt
Christensen,
TSRC,
UC
Berkeley
Ø Special
thanks
to
the
worldwide
shared
mobility
operators
and
experts
who
make
our
research
possible,
including
Timothy
Papandreou
and
Russell
Meddin

24. www.tsrc.berkeley.edu
Email:
sshaheen@berkeley.edu
Twitter:
SusanShaheen1