Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 107
Car-Sharing: Where and How It Succeeds
the greater impacts are due to decreasing vehicle:member ratios, rather than
to a greater proportion of members selling their cars.
Another, simpler way of examining changes in vehicle ownership is to ask
members what they would do if car-sharing services stopped. These results
are discussed in full below, but it should be noted here that nearly one-third
stated that they would buy a car.
4.3Travel Behavior Changes and Related Impacts
Potential Impacts
Reduced Vehicle Travel
Car-sharing, according to its proponents, can have a major impact on the
travel behavior of its members by reducing the number and length of trips.
This is largely a function of changes in vehicle ownership: once members
give up their cars, the automobile will no longer be the "default mode" for
every trip. Rather, it is argued, members will weigh up the cost, travel time
and comfort of different modes of travel, such as transit, car-sharing and
walking, on a more rational basis, before deciding which to use for a par-
ticular trip. In turn, reduced vehicle travel translates into a range of other
benefits some straightforward, such as reduced emissions, and some more
speculative, such as increased physical activity and support for local shops
and services.
The manner in which car-sharing converts fixed driving costs into variable
ones may be largely responsible for these changes in travel behavior. The
costs of driving can be divided into fixed costs, such as car payments and
insurance, and variable costs, such as gasoline, tolls and non-residential
parking. Once the decision has been made to own a car, these fixed costs
are (correctly) treated as sunk costs by a household (Steininger, Vogl &
Zettl, 1996), and perceptions of the cost of a trip are based on variable costs
or even just gasoline and parking alone. Since fixed costs account for
the majority of driving costs 80% or more for a car that is driven 10,000
miles per year or less this means that the economics of driving are heavily
skewed (Exhibit 4-7).
Most car-sharing operators, in contrast, charge for time used and/or mile-
age driven, meaning that almost all driving costs become variable and are
highly visible to members. In the perceptions of members, driving therefore
becomes more expensive, and car-sharing members, "mindful of the cumu-
lative costs of driving," practice a "more resourceful form of automobility"
Page
4-13
OCR for page 108
Chapter 4 · Impacts of Car-Sharing
(Cervero & Tsai, 2003). As shown in Exhibit 4-8, there is a striking contrast
between the marginal costs of a trip for car-sharing and private vehicle
ownership. It suggests that the threshold for the cost-effectiveness of car-
sharing is 5,000 miles per year.5
Exhibit 4-7 Components of Driving Costs
Variable (Operating) Costs Cost Per Mile
Gas and oil $0.061
Maintenance $0.039
Tires $0.015
Added depreciation (per 1,000 miles above 15,000/yr) $161
Fixed (Ownership) Costs Cost Per Year
Insurance $1,181
License, registration, taxes $167
Depreciation (15,000 miles/yr) $3,051
Finance charges $554
Miles Driven per Year
Annual Costs 10,000 15,000 20,000
Fixed Costs $4,953 $4,953 $4,953
Variable Costs $1,150 $1,725 $3,105
Total Costs $6,103 $6,678 $8,058
Fixed costs as a % of total costs 81% 74% 61%
Source: Adapted from American Automobile Association (2003).
Figures are for a small car (2003 Chevrolet Cavalier LS).
5. See Section 4.4 for a discussion of the applicability of AAA motoring cost data.
Page
September 2005
4-14
OCR for page 109
Car-Sharing: Where and How It Succeeds
Exhibit 4-8 Cost per Trip, Car-Sharing vs. Ownership
$17,500
Motoring costs based on
Car-Sharing
$15,000
Car Ownership
AAA figures. Car-sharing
costs based on I-GO
$12,500 (Chicago) July 2004 rates,
assuming hourly costs
Annual Motoring Costs
equate to $0.60 per mile,
$10,000
i.e. members drive 10
miles per hour reserved.
$7,500 Actual costs will be lower
if rental cars are used for
$5,000 longer trips.
$2,500
$0
0
2,5
5,0
7,5
10
12
15
,00
,50
,00
00
00
00
0
0
0
Miles Driven Per Year
While changes in vehicle ownership may be the main driver behind reduced
vehicle travel, proponents have also suggested other mechanisms through
which car-sharing can produce these benefits:
· Provide access to a car during the working day. Many employees
drive to work because they need a car during the working day
for example, to visit clients or run errands. According to surveys
in the San Francisco Bay Area, 11% of commuters cite the need for
a car for work as an impediment to commuting by transit, bicycle
or carpool (RIDES for Bay Area Commuters, 2003). Providing
car-sharing at workplaces may help to eliminate this barrier, and
many employers have introduced car-sharing as a part of their
commute trip reduction program.
· Promote compact development. As discussed in the section
above, car-sharing can help to increase development densities
through lowering parking ratios. In turn, a large body of research
suggests that compact development reduces vehicle travel as
residential density doubles, vehicle miles traveled per capita falls
by approximately 20% (Holtzclaw et al., 2002). While this mecha-
nism is perhaps the most difficult to confirm directly through
empirical studies, it may represent the most important long-term
benefit if car-sharing is introduced on a large scale.
· Provide mobility insurance. "Guaranteed Ride Home" programs
have had considerable success in persuading people to carpool
or ride transit to work, through providing a guarantee that they
will not be stranded in the event of an emergency such as a sick
Page
4-15
OCR for page 110
Chapter 4 · Impacts of Car-Sharing
child, or needing to work late. Car-sharing may function in the
same way, if cars are provided close to the workplace or at transit
stations.
Induced Vehicle Travel
At the same time as car-sharing may reduce vehicle travel among some
members, it provides easier access to a vehicle for members who did not
previously own a car. While some car-sharing trips may have been otherwise
made by rental or borrowed cars, or by taxi, others are likely to represent
new vehicle trips. Members may use car-sharing to access new destinations,
or substitute for trips previously made by transit, bicycle or walking.
It is important to recognize that induced travel can be viewed as a benefit, if
car-sharing is serving to promote greater mobility and reduce travel times
for members. One example often cited by car-sharing proponents concerns
lower-income households, where providing access to a car may help to
overcome issues of social exclusion. Another, in the context of campus car-
sharing, relates to providing mobility for students who may be subject to
restrictions on bringing their own cars to school. However, induced travel
will offset or reverse any reductions in vehicle travel from members who
drive less.
Transit Ridership
Changes in transit ridership will largely depend on the net impacts on vehicle
travel. Should overall vehicle travel fall, i.e. with reduced travel outweighing
induced travel, this is likely to be realized partly as an increase in transit
ridership, along with greater walking and cycling. Should vehicle travel rise,
car-sharing may substitute for some trips formerly made by transit.
There are also three potential impacts on transit ridership specifically:
· Combined transit/car-sharing trips. Many car-sharing operators
point to the potential to take transit for the line-haul segment of
the journey, before picking up a car-sharing vehicle at the station
to travel the "last mile" to the destination. This may be particular-
ly important to reach suburban locations from congested central
cities, or where there are bridge tolls or similar psychological bar-
riers. For example, a Manhattan resident might take Metro North
rail service to White Plains, NY, and pick up a car there to drive
the final leg of the journey.
· Access to car-sharing vehicles. In some cases, members may
not live within easy walking distance of a car-sharing location, or
vehicles at the closest location may already be reserved. In these
cases, they may use transit as an access mode.
Page
September 2005
4-16
OCR for page 111
Car-Sharing: Where and How It Succeeds
· Changes in peak orientation. Most transit trips, especially those
that utilize the park-and-ride facilities at transit stops, are heav-
ily peak oriented, placing demands on the system at a time when
capacity is most constrained (see, for example, Chesapeake Bay
Foundation, 2001). In contrast, car-sharing trips often take place at
evenings and weekends, when surplus capacity is available. This
has particular implications where station lots regularly fill to ca-
pacity, and where park-and-ride commuter spaces are given over
to car-sharing. In this instance, peak commuter trips are likely to
be replaced by off-peak transit-car-sharing trips.
Emissions and Gasoline Consumption
As with transit ridership, the impacts of car-sharing on emissions and gaso-
line consumption will largely depend on its net impact on vehicle travel.
However, there may be additional benefits from the use of more fuel-efficient
cars by car-sharing operators. Even if vehicle travel remains unchanged, re-
ductions in emissions and fuel consumption could therefore be expected.
These additional benefits may be realized in three main ways:
· Use of alternative-fuel vehicles. Hybrids and electric vehicles
have been used by many car-sharing operators (Chapter 2), and
some automobile manufacturers have seen this as a way to meet
mandates for the introduction of low-emission vehicles (Shaheen,
Wright & Sperling, 2002). However, they account for a relatively
small proportion of the car-sharing fleet at present.
· Replacement of older vehicles. It has been suggested that car-
sharing members tend to own disproportionately older, more pol-
luting vehicles. To the extent that these are given up as members
join the program, car-sharing will bring further emissions benefits.
(See, for example, Rydén & Morin, 2005.)
· Use of smaller cars. The core fleet of most car-sharing operators
consists of compact, fuel-efficient cars such as a Scion or a Honda
Civic. Larger station wagons, sport utility vehicles and pick-up
trucks are available for special purposes (Chapter 2). This means
that members have the ability to select the "right" vehicle for a
specific purpose, rather than using a large household car for all
trips, from errands to camping expeditions. At present, however,
this impact is more speculative compared to the benefits from
alternative-fuel and newer, fuel-efficient vehicles, for which there
are firmer data (see following section).
Exhibit 4-9 shows some comparisons among vehicles commonly used by
car-sharing operators, indicating that use of hybrid vehicles can more than
halve gasoline consumption and reduce smog-forming emissions.
Page
4-17
OCR for page 112
Chapter 4 · Impacts of Car-Sharing
Exhibit 4-9 Fuel Economy and Emissions Ratings of Vehicles
Commonly Used by Car-Sharing Operators
Vehicle City MPG EPA Air Pollution Score*
Toyota Prius (hybrid) 60 8-9
Honda Civic 29-36 2-6
Scion xA and xB 31-32 2-3
Dodge Neon 25 3
Volkswagen Jetta 22-24 3-9
Ford Explorer 2WD 15-16 1-6
* Air Pollution Score refers to pounds of smog-forming emissions per 15,000 miles. A rating of 10
indicates the lowest emissions (0-1.0 lb).
Source: US EPA (www.fueleconomy.gov), figures for 2005 models.
Empirical Evidence
Methodological Considerations
In general, the empirical evidence on changes in travel behavior is less
definitive compared to that concerning vehicle ownership impacts. This is
partly a reflection of complex methodological problems. Surveys offer the
simplest approach, but self-reported data on vehicle mileage particularly
for years past is unlikely to be reliable (Katzev, 2002). Travel diaries tend
to only cover a short time period for a small sample. While mileage driven
in car-sharing vehicles is readily available from most reservations systems,
this does not include mileage driven in rental cars, borrowed cars, or other
household vehicles. Research also runs into questions of causation. While
members may drive less once they join a car-sharing program, is this due
to car-sharing itself, or was their decision to join a reflection of external fac-
tors causing them to reduce their vehicle travel, such as moving to a more
urban area?
An alternative, more sophisticated approach, employed by Cervero & Tsai
(2003), uses travel diaries coupled with a control group methodology. How-
ever, while potentially avoiding problems of inference, this approach does
bring its own set of problems. In the Cervero research, weekdays/weekends
and work/non-work days were analyzed separately. However, the only sta-
tistically significant change in vehicle travel was obtained for weekday work
days whereas any induced travel would be expected to be on weekends,
when shared cars are used the most.
It is extremely difficult to find a good control group. In the Cervero research,
Page
September 2005
4-18
OCR for page 113
Car-Sharing: Where and How It Succeeds
the control group consists of members who had expressed interest in joining
a car-sharing program, but had yet to do so, for example because service
was not available in their neighborhood. However, subsequent expansions
mean that car-sharing is now available to most San Francisco residents, sug-
gesting that there are other factors behind the reason not to join. There are
major differences between the sample of members and the control group,
such as differing levels of education, car ownership, and baseline vehicle
travel. As Lane (2005) points out: "These differences imply differing mobility
needs that may have contributed to (1) contrasting decisions of whether to
join City CarShare, or (2) affected changes in travel behavior independent
of membership."
Changes in Vehicle Travel
Exhibit 4-10 summarizes published studies on the impacts of car-sharing on
vehicle travel. As can be seen, many studies show no statistically significant
change. However, those that do suggest that car-sharing does lead to reduced
vehicle travel, although the magnitude of the change varies considerably.
The key variable is often the relative proportion of members who gave up
a vehicle who will usually drive less as a result and members who gain
access to a car, who will account for most of the induced travel. It should be
noted that the magnitude of the former will generally outweigh that of the
latter. As Lane (2005) points out in the Philadelphia context, "any upward
movement in the miles driven by members who gained access to a car is
limited to the small amount they currently drive PhillyCarShare's vehicles,
which averages just 29.9 miles per month (33.9 for members who avoided
acquiring a car, and 27.5 miles for members who simply gained access)."
There is also evidence that the impacts on vehicle travel change over time, as
a car-sharing program matures. An evaluation of the City CarShare program
in San Francisco found a net increase in vehicle travel after the first year, but
a reduction after two years. Cervero & Tsai (2003) conclude:
Evidence of travel suppression stands in stark contrast to first-year
impacts wherein members' average VMT had increased. Early adopters,
many drawn from the ranks of environmentalists and avid cyclists who
owned no car, began logging vehicle miles on the streets of San Francisco;
with time, as the program has attracted a more mainstream clientele, the
novelty of car-sharing has worn off, and members have shed car ownership,
"induced travel" appears to have been replaced by "reduced travel."
Page
4-19
OCR for page 114
Chapter 4 · Impacts of Car-Sharing
Both the San Francisco and Portland studies also suggest that vehicle travel
per member declines over time possibly as they become more aware of the
real cost of each vehicle trip, or as the "novelty effect" of car-sharing wears
off over time (Katzev, Brook & Nice, 2000; Cervero & Tsai, 2003).
There is even less information on the cost-effectiveness of car-sharing as a
trip reduction strategy, given the difficulties in calculating both the public
investment and the total reduction in vehicle travel. The results of these
types of calculations will depend on the way in which "public investment"
is calculated, since many programs have received a combination of financial
and in-kind support from several different agencies. One study suggests that
car-sharing is relatively expensive compared to other demand management
programs, but that the costs should decline as operators become more self-
sufficient. Using British data, Cairns et. al. (2004) estimate the cost at about
15 cents per mile of travel reduced.6 On the other hand, communities that
have offered limited public support for car-sharing have received the trip
reduction benefits essentially for free.
It is important to note that studies have focused on the impacts of residential
car-sharing. Only one study reported here has considered business car-shar-
ing in detail, and the overall effects were unclear.
6. Using April 2005 exchange rate of $1.885 US to one British pound.
Page
September 2005
4-20
OCR for page 115
Car-Sharing: Where and How It Succeeds
Exhibit 4-10 Research on Vehicle Travel Impacts
Context Results Reference
European Research
Germany The share of travel accounted for by motor vehicles fell from 63% to Baum & Pesch (1994), cited
41%, after members joined. in Harms & Truffer (1998)
Austria For users who previously had a car available, annual VMT fell by 62%, Steininger, Vogl & Zettl
from 10,088 to 3,855 miles. For users with no car previously available, (1996)
annual VMT rose by 118%, from 830 to 1,809 miles.
Netherlands Reduction in VMT for both former car owners (37% reduction) and mem- Meijkamp & Theunissen
bers who did not previously own a car (29%). (1996), cited in Lane (2005)
Switzerland Members who give up their cars after joining reduce their annual motor Muheim & Partner (1998)
vehicle travel from 5,779 to 1,616 miles per year a reduction of 72%
(or 58% if travel by motorcycle is included).
Bremen, Average VMT fell by 32% after members joined, from 3,144 miles per Koch (2002)
Germany year to 2,133. Figures refer to those with a combined car-sharing/transit
annual pass.
United Kingdom Members who previously owned a car (51% of the sample) reduced VMT Ledbury (2004)
by 1,123 miles per year (26%). Members who did not previously own a
car increased annual VMT by 473 miles. Note small sample size (n=96).
Bremen, Average VMT reduction of 1,925 miles per year (45%), leading to a 54% Rydén & Morin (2005)
Germany reduction in CO2 emissions*. This includes a small group of members (7%
of the sample) who increased annual VMT by 500 miles (60%).
Belgium Average VMT reduction of 1,865 miles per year (28%), leading to a 39% Rydén & Morin (2005)
reduction in CO2 emissions*. This includes a small group of members (7%
of the sample) who increased annual VMT by 745 miles (14%). Most
respondents (65%) reported no change.
Bremen, Germany Business car-sharing may lead to a slight increase in total car mileage for Rydén & Morin (2005)
and Stockholm, work-related purposes, given easier access to vehicles. However, nearly
Sweden 30% of employees report that car-sharing has helped them drive to work
less often.
North American Research
Portland, OR Vehicle mileage for those who owned a vehicle at time of joining fell by Katzev, Brook & Nice
18%, from 103.3 to 84.4 miles for the week of the trip diary, although (2000)
this result was not statistically significant. For those who did not own a
vehicle, it rose from 0.3 to 24.9 miles.
Portland, OR Travel diaries suggested a 7.5% reduction in vehicle travel after joining, Cooper, Howes & Mye
but results not statistically significant. (2000)
San Francisco, Members' daily VMT (weekday/workday) fell from 2.80 to 1.49** miles. Cervero & Tsai (2003)
CA Among the control group of non-members, it rose from 5.45 to 20.85.
These figures refer to second-year impacts; first-year impacts showed a
net increase in VMT.
Arlington, VA Members reported a reduction in VMT of 43%, or 3,250 miles per year. Price & Hamilton (2005)
45% of respondents reported reducing driving after joining car-sharing,
while 35% said they increased their driving. Nearly half of respondents
(49%) said they walk more often because of their involvement in car-
sharing, and 54% said they use transit more often.
* Based on changes in vehicle fleet mix, since car-sharing vehicles tend to be more fuel-efficient. Includes
increased emissions from public transportation.
** This measure is referred to as "mode adjusted VMT" by Cervero & Tsai.
Page
4-21
OCR for page 116
Chapter 4 · Impacts of Car-Sharing
The results of our internet survey also show a substantial decline in the
number of miles driven. (It should be noted that since aggregate annual
figures were requested and that travel diaries were not used, the accuracy
of the figures reported may be less than ideal.) Respondents, on average,
reported driving about 3,850 miles per year (320 miles per month) at the
time of the internet survey. This is approximately 63% of the average mile-
age that they previously drove, which is a substantial reduction in driving
of almost 40% (Section 3.1). Note that Zipcar, which did not participate in
the survey, claims an even greater reduction based on its own 2004 survey
an almost 80% reduction in VMT from 5,295 to 1,068 miles per year after
joining car-sharing (Zipcar, 2005).
While, on balance, there was a substantial net reduction and more than 45%
of households reported driving less, many households did increase their
VMT substantially (Exhibit 4-11). Since there may have been many recent
changes in the lives of these members (employment, marriage, home loca-
tion, etc.), not all of the changes in vehicle miles should be attributed to their
participation in car-sharing.
Exhibit 4-11 Self-Reported Changes in Vehicle Travel
Change in VMT After Percent of Respondents
Joining Car-Sharing
Reduced Travel
By less than 50% 18.3%
By more than 50% 27.5%
No change reported 28.6%
Increased vehicle travel
By less than 50% 11.5%
By more than 50% 14.7%
Figures refer to respondents who provided both before and after information.
Source: Car-Sharing Member Survey.
Changes in VMT and percentage changes in VMT did show some variation
according to specific characteristics. In terms of changes in vehicle miles
traveled:
· Auto ownership: There was a bi-modal split among the respon-
dents who did not currently own cars, meaning that more than
average numbers of respondents reported either a substantial de-
crease in mileage (more than 7,500 miles per year) or a substantial
increase in mileage (more than 2,500 miles per year).
Page
September 2005
4-22
OCR for page 117
Car-Sharing: Where and How It Succeeds
· Age group: Another bi-modal split was in evidence among per-
sons in the 25 to 34 year old group, who reported either a substan-
tial decrease in mileage (a decrease of 5,000 miles or more) or a
substantial increase in mileage (more than 2,500 miles per year).
· Gender: Another bi-modal split: females reported either a sub-
stantial decrease in mileage (a decrease of 5,000 miles or more) or
a slight increase in mileage (from 1 to 2,500 miles more).
· Education: Persons with a Bachelor's degree more often than
average reported a small increase in mileage (from 1 mile to 2,500
miles more).
· Operator: Members of one car-sharing company were more likely
to report decreases of 2,500 to 10,000 miles than the members of
other car-sharing companies, whose members tended to show no
larger or smaller than average figures for any of the changes in
mileage.
· There were no substantial variations evident due to income or
household size.
In terms of percentage changes in vehicle miles traveled:
· Auto ownership: There was a bi-modal split among the respon-
dents who did not currently own cars, meaning that more than
average numbers of respondents reported either a substantial
percentage decrease in mileage (76% or more) or a substantial
percentage increase in mileage (more than 26%).
· Age group: Persons in the 25 to 34 year old group more often than
average reported a 26% to 75% increase in mileage.
· Gender: Females more often than average reported a percentage
increase in mileage of 50% or more.
· Education: Persons with a Bachelor's degree more often than
average reported a small increase in mileage (from 1 mile to 2,500
miles more).
· Household size: Persons in two-person households were more
likely than average to report mileage decreases of from 1% to 50%.
· Operator: Members of one car-sharing company were more likely
to report decreases of 26% or more than the members of other car-
sharing companies, whose members tended to show no larger or
smaller than average figures for any of the percentage changes in
mileage.
· There were no substantial variations evident in percentage chang-
es in mileage due to income or education.
Page
4-23
OCR for page 118
Chapter 4 · Impacts of Car-Sharing
There are complex travel effects for those persons who, as a result of joining
car-sharing, sold their car or the family car. Of those who sold a car and
provided information on changes in vehicle miles traveled, 36.9% decreased
their miles traveled, 28.6% made no change, and 34.5% increased their VMT.
More than half of those persons who sold cars and increased their mileage
driven (56.7%) had greater than 50% increases in their miles driven.
Other Travel Behavior Impacts
While a relatively small number of studies have examined changes in vehicle
miles traveled which generally requires the use of travel diaries others
have gathered more qualitative information on changes in travel behavior.
Typically, questions ask whether the respondent walks, cycles or takes transit
more or less after joining. Exhibit 4-12 shows an example from Philadelphia,
which is fairly typical of results obtained elsewhere.
Exhibit 4-12 Self-Reported Changes in Travel Behavior, PhillyCar-
Share Members
90%
80% 77% More Since Joining
Less Since Joining
7 0%
6 0%
% of Member Subset
50% 48 %
4 0% 37% 37% 37%
30% 27%
1 9%
2 0%
14%
12%
1 0% 8%
6% 6% 6%
3% 4%
2% 2% 2% 2%
0%
0%
Walk Bike Transit Taxi Drive Walk Bike Transit Taxi D r i ve
Members Who Reduced Their Car Ownership Members With No Car Before Joining
Source: Lane (2005)
Other surveys have probed the issue of mode of access to car-sharing
important if the effects on transit ridership are to be quantified. In San
Francisco, 68% of City CarShare trips were accessed on foot, 18% by tran-
sit, and 9% by bicycle (Cervero & Tsai, 2003). City CarShare also reports
considerable success in encouraging its members to make combined tran-
sit/car-sharing trips, particularly to avoid congestion and tolls on the San
Page
September 2005
4-24
OCR for page 119
Car-Sharing: Where and How It Succeeds
Francisco-Oakland Bay Bridge. For example, 25% of trips using the City
CarShare vehicle at Ashby BART station in Berkeley are taken by members
with a San Francisco home address, compared to 2-3% at other Berkeley
and Oakland locations.7
In Germany and Belgium, Rydén & Morin (2005) report that the net increase
in transit ridership is 685 miles per member per year, with the largest in-
crease occurring on weekends. They find that 5% of members ride transit
less, 22-32% ride more, and 63-72% report no change. Muheim & Partner
(1998) reports that most of the increased transit demand from car-sharing
in Switzerland takes place at off-peak times.
There is also empirical support for the theory that car-sharing promotes
the use of more fuel-efficient vehicles. Cervero & Tsai (2003), for example,
found that vehicles owned by car-sharing members in San Francisco tended
to be older on average nine years old, with an odometer reading of 73,000
miles. They conclude that City CarShare reduces gasoline consumption and
emissions, partly because of reduced automobile travel, but also because
car-sharing vehicles tend to be small, fuel-efficient and carry several people.
Rydén & Morin (2005) conclude that car-sharing vehicles consume 11% less
fuel on average, compared to the vehicles given up by members.
Results from our internet survey support many of the positive travel behavior
outcomes of car-sharing noted in other studies, in terms of both a reduc-
tion in vehicle travel and an increase in overall mobility. These qualitative
outcomes are shown in Exhibit 4-13.
Respondents reported being able to get to destinations that were formerly
not accessible and to travel more often. One of the focus group respondents
reported that "I'm less reluctant to go to suburban parties now because I used
to have to stay overnight or get a ride back from a drunk stranger now I
can zip out and zip back and it's so easy." Respondents also reported that
they made more multi-purpose trips (i.e. trip chaining) and used transit more
often. Overall, 83.1% of the respondents said that they felt safe when using
public transit: 54% agreed with this statement and 29.1% strongly agreed.
On the other hand, several focus group members reported that they take
fewer longer-distance local trips with car-sharing than with an owned car
because of the costs associated with higher mileage trips.
Exhibit 4-13 also provides specific results for respondents who reported
7. Analysis of 2004 City CarShare data by Bryce Nesbitt.
Page
4-25
OCR for page 120
Chapter 4 · Impacts of Car-Sharing
selling a car due to joining car-sharing. The most notable differences are
that persons who sold a car more often reported driving less and saving
money on transportation.
Focusing on impacts, Exhibit 4-14 adds to demographic information from
the internet survey presented in Chapter 3 by indicating how the impacts
of car-sharing differ by demographic groups. Age seems to make the most
difference of all the demographic variables, having particularly strong
relationships in terms of making fewer trips by auto, using transit more
often, and walking more often. Income levels come into play in a number
of instances, but education does not, which is somewhat surprising.
Page
September 2005
4-26
OCR for page 121
Car-Sharing: Where and How It Succeeds
Exhibit 4-13 Effects of Car-Sharing Membership on Travel Behavior
Neither
Strongly agree or Strongly
Effects of involvement in car-sharing Members disagree Disagree disagree Agree agree
Make fewer trips by auto All 6.2% 10.0% 14.9% 35.3% 33.7%
Gave up car 2.3% 4.2% 9.2% 36.4% 47.8%
Save money on transportation All 6.3% 14.9% 16.7% 33.1% 29.0%
Gave up car 3.5% 13.5% 14.7% 26.8% 41.5%
Able to get to places I couldn't All 8.4% 6.0% 26.7% 26.3% 32.6%
Gave up car 5.8% 5.1% 26.9% 27.6% 34.6%
Able to travel more often All 9.1% 20.8% 25.4% 23.3% 21.4%
Gave up car 9.1% 24.3% 28.4% 21.3% 16.9%
Use transit more often All 9.7% 22.1% 28.6% 23.1% 16.6%
Gave up car 8.7% 27.1% 31.0% 20.5% 12.7%
Walk more often All 10.3% 18.1% 34.3% 22.6% 14.7%
Gave up car 9.0% 21.8% 37.4% 20.9% 10.9%
Can make more multi-purpose trips All 20.6% 12.1% 45.0% 8.3% 14.1%
Gave up car 19.6% 13.5% 50.4% 6.4% 10.1%
Source: Car-Sharing Member Survey.
Page
4-27
OCR for page 122
Chapter 4 · Impacts of Car-Sharing
Exhibit 4-14 Reported Demographics of Car-Sharing Companies
Effects of car-sharing Demographic characteristics more frequent
to this group than to all respondents in
general
Make fewer trips by auto Ages 34 and younger; not 55 and older *
Females
Not car owners
Incomes less than $20,000 and not over
$125,000
Members of one Canadian company more often
strongly agreed
Save money on transportation Males
Age 55 to 64
Members of one Canadian company more
frequently strongly agreed; members of one
US company more often disagreed or strongly
disagreed
Able to get to places I couldn't Not ages 24 and under
Females
Incomes of $50,000 to $60,000 and $100,000
to $125,000, but not less than $30,000
Members of one Canadian company more often
strongly agreed
Able to travel more often Age 55 and over; not ages 24 and under
Males
Incomes $10,000 or less
Use transit more often 45 and older; not 24 and under *
Members of one US and one Canadian company
more often agreed or strongly agreed
Walk more often 55 and older *
Members of one US and one Canadian company
more often agreed or strongly agreed; members
of one US company more often disagreed
Can make more multi-purpose trips Age 65 and over; not ages 24 and under
Not car owners
Incomes not less than $20,000
Members of one US company split their an-
swers: they more often than average disagreed
or agreed than members of other companies
* This is an especially strong relationship.
Source: Car-Sharing Member Survey.
Page
September 2005
4-28
OCR for page 123
Car-Sharing: Where and How It Succeeds
As noted in Chapter 3, having access to an automobile was seen as a distinct
advantage by many car-sharing members, since there are many destinations
and activities in today's world that are much more accessible by cars than
by any other mode. Respondents reported that car-sharing was particu-
larly important to them for recreation and social trips, other shopping, and
grocery shopping. Special activities like transporting family and friends,
moving furniture or hauling large loads, going to medical appointments,
and visiting relatives also were made much easier and more comfortable
when using a car.
Effects on travel behavior would not be complete without an assessment
of how car-sharing members got to those cars. The overwhelming major-
ity walked to the car-sharing vehicle (Exhibit 4-15) unsurprisingly, since
most members have car-sharing within a few blocks of both their home and
workplace (Exhibit 4-16).
Exhibit 4-15 Mode of Access to Car-Sharing
Mode of Access Percent of Respondents*
Walk 75.2%
Public transportation 18.6%
Bicycle 8.7%
Passenger in car 2.6%
Taxi 0.9%
Other 1.7%
* Multiple responses permitted.
Exhibit 4-16 Distance to Car-Sharing
Distance to Car-Sharing Location From Home From Work
At work N/A 9.0%
Within a block 28.0% 23.2%
Within several blocks 57.2% 47.2%
Need to take transit/other vehicle 14.7% 20.3%
The converse of information about changes in travel behavior due to car-
sharing is the speculation by car-sharing members about how their lives
would be different if they did not have access to car-sharing. (Some of this
information is also discussed in Chapter 3.) If car-sharing had not been
available for the last trip taken using car-sharing, 29.3% of the respondents
to the internet survey would not have made that trip; another 20% would
Page
4-29
OCR for page 124
Chapter 4 · Impacts of Car-Sharing
have used public transportation. If car-sharing services stopped, respondents
reported that they would take a number of actions8, including using public
transit more often (more than one-third of the respondents), while about
one-third reported that they would get rides from friends, one-third would
use taxis more often, and just less than one-third would buy a car.
Overall, these results suggest some complex impacts on transit ridership.
On the one hand, car-sharing is clearly substituting for many transit trips.
Members are choosing car-sharing because they had things to carry, for
example, or had to make multiple stops (see Chapter 3); while feasible by
transit, these trips may be time-consuming or difficult. This substitution
effect, however, is likely to be somewhat less than the 20% who stated that
they would have used public transportation for their last trip had car-shar-
ing not been available, because many of these respondents have also given
up a car since joining car-sharing and would presumably have used their
own vehicles instead. Moreover, any substitution effect is limited by the
relatively infrequent use of car-sharing members in our survey report a
median of two trips per month (Section 3.1).
On the other hand, many new transit trips are being generated. Partly this
is due to the use of transit as an access mode, which (at 19% of car-sharing
trips) is likely to cancel out any substitution effect from use of car-sharing
instead of transit. On top of this, there is evidence of a wider increase in
transit ridership, most probably resulting from a reduction in vehicle own-
ership and workplace Transportation Demand Management programs.
Nearly 40% of members state that they use transit more often as a result of
their involvement in car-sharing.
As well as using car-sharing, 81.2% of the respondents also rent cars from
time to time, for the multiple reasons shown below:
· Rented cars cost less for longer trips 84.5%
· Don't have to worry about mileage charges 44.0%
· Have more types of vehicles with rentals 15.1%
· Other reasons 19.3%
8. Multiple responses permitted.
Page
September 2005
4-30
