A REVIEW OF THE CHANGING PRICES AND TAX LEVELS FOR

NEIGHBORHOOD CARSHARING IN THE UNITED STATES:

2011 - 2016

Paper for Presentation at the 2017 Transportation Research Board Annual Conference

November 6, 2016

Joseph P. Schwieterman, Ph.D.

Professor, School of Public Service

Director, Chaddick Institute for Metropolitan Development

DePaul University,

14 E. Jackson, Suite 1600

Chicago, IL 60604

312.362.5732 (voice) 312.362.5506 (fax)

jschwiet@depaul.edu

1

ABSTRACT

This study explores the changing prices and level of taxation for neighborhood carsharing

through a review of hourly rates in 80 U.S. cities and an analysis of the pricing by three major

carsharing companies: car2go, Maven, and Zipcar. The results show that, between 2011 and

2016, the base price of a one-hour Zipcar carshare dropped from $9.39 to $8.92, a 5.0% decline.

In real (inflation-adjusted) terms, prices experienced an even a more dramatic 11.2% reduction,

suggesting that the sector is facing significant downward price pressure. Each of the three

operators evaluated, car2go Maven, and Zipcar, all have the lowest prices in some scenarios.

Retail taxes, which were 15.5% in 2011, rose to 17.0% in 2015, offsetting almost a third of the

price reduction. Interpreted broadly, these results support the notion that carsharing is a

maturing industry characterized by extensive competition and retail tax rates that are

substantially higher than most other segments of the “sharing economy.”

2

I. INTRODUCTION

A key consideration for many consumers when deciding whether to join a carsharing

organization is the extent to which using this mobility option can reduce their expenditures on

transportation. However, despite the existence of an expansive body of research, relatively little

publically available data exists on the prices and taxes paid for carsharing, or how these costs

have changed over time.

This paper attempts to partially fill this void by reviewing the prices charged for hourly

reservations at three major providers: car2go, Maven, and Zipcar, and the taxes imposed on this

sector. The study explores how both rates and taxes changed from 2011 to 2016 to better

understand this sector’s evolving competitive position. This study also summarizes the results of

a prediction model showing how hourly prices vary between firms and in accordance with a

city’s size.

Pricing information was obtained by recording the base prices and applicable taxes at

locations in 80 cities throughout the United States. The data set assembled includes the country’s

largest cities in which neighborhood carsharing is available, as well as a stratified sample of an

equal number of smaller cities. To ascertain the prevailing rate of tax, data analysis concentrated

on the price quotes provided for reservations of various lengths on the official website for Zipcar

(zipcar.com), which is the only major provider that makes this information readily available to its

members prior to a purchase.

The results are provided in six sections. Section II offers a background perspective on the

growth of the industry and prior literature exploring the carsharing market. Section III describes

the methods used to measure the cost of carsharing. A summary of the results, starting with an

evaluation of changes in the hourly rates charged (Section V), and concluding with the prevailing

taxes paid by the sector, is then provided. The final section (VI) offers conclusions and

recommendation for further study.

II. BACKGROUND AND RELEVANT LITERATURE

Carsharing has evolved greatly since Car Sharing Portland, the country’s first large-scale

program, emerged on the West Coast in 1998. This Oregon company, like most of the other

early entrants, provided services similar to more established carsharing organizations that existed

in Canada and Europe at the time. These pioneering operators emphasized a neighborhood

residential model and focused on pods strategically scattered throughout dense urban areas.

Gradually, carsharing became prevalent in smaller cities and suburbs as well as specialty

locations, including government facilities, residential complexes, and airports. Dozens of

universities also now have active carsharing pods.

3

Although some of the characteristics of carsharing have changed, it still generally works on

a membership-based model, which includes automatic insurance coverage and typically entails

an annual fee. Unlike most car rentals, carsharers do not enter a separate contract every time they

use a vehicle. Carsharing also involves filling up the tank when fuel runs low (many providers,

including Zipcar, keep a credit card in vehicles to make fill-ups relatively easy). Members are

also expected to clean up after themselves or membership may be revoked.

With this sector’s growth, there has come extensive research about its environmental,

economic, and social benefits (10, 11 & 13). This research points to the reductions in emissions,

pollution, congestion, and parking requirements made possible by this sector—much of it

stemming heavily from the reduction in privately owned vehicles in urban settings. Carsharing

promotes active lifestyles by increasing walking and biking without posing significantly

downward effects to transit use. This is partially due to the fact that many carsharing members

simultaneously increase their reliance on buses and trains (7, 9, 12 & 14). The reduction in

demand for parking spaces afforded by carsharing also increases open space and public safety,

and boosts local economies.

Carsharing encompasses both for-profit businesses and non-profit businesses. Zipcar is

by far the country’s largest for-profit provider, with pods in more than 50 U.S. cities (18). The

Boston-based company, which went public in 2011 before being acquired by Avis Company in

2013, serves 38 U.S. states as well as various locations in Canada and Europe, and has reported

having more than 10,000 vehicles in its worldwide fleet (17). Other providers—both private and

nonprofits—include Buffalo CarShare (Buffalo, NY), City CarShare (San Francisco, CA), eGo

CarShare (Denver, CO), Enterprise CarShare (multiple cities), Getaround, Hourcar (Chicago,

IL), and CarHoppers (multiple cites).

Since 2015, automobile manufacturers have made significant inroads into carsharing.

Audi, BMW, Daimler AG, General Motors, and Ford all now have footholds in certain U.S.

cities, although none are as large in the United States as Zipcar (Table 1) (6). Their presence is

significant not only for the heightened competition it provides, but also for innovative pricing

strategies, such as sophisticated pay-per-minute pricing and differential fees for time spent

driving versus parked (in the case of BMW’s ReachNow), being employed. Auto manufacturers

are interested in carsharing as an incremental step toward preparing for the dramatic changes in

vehicular travel that could occur with the widespread deployment of autonomous vehicles (6).

4

TABLE 1 Notable U.S. Carsharing Services Affiliated with Automobile Manufacturers

Service Launch Year

and Description Geographic Scope Pricing Model

AUDI

at home

Launched in late 2015;

focus on luxury sedans

Miami, FL,

San Francisco, CA

Hourly or daily fees

with roundtrip focus

BMW’s

Reach Now

Launched in early 2016 Seattle, WA Normally

$0.41/minute driving

and $0.30 when

parked; focus on one-

way trips

DAIMLER AG’s

car2go

Launched in the U.S. in

2011; now the world’s

largest carsharing

service

Austin, TX

Columbus, OH

Denver, CO

Minneapolis & St. Paul,

MN

San Diego, CA

Portland, OR

Seattle, WA

Washington, DC

Normally

$0.41/minute; focus

on one-way trips

FORD

GoDrive

Launched in mid-2015 London, England

(U.S. expansion

anticipated)

Priced per minute;

focus on one-way trips

GENERAL

MOTORS Maven

Launched in early 2016;

now available in five

U.S. cities

Boston, MA

Chicago, IL

Detroit, MI

Washington, DC

$8.00/hour in many

settings; roundtrip

focus

Source: Websites and press releases of companies

New services, most notably one-way and peer-to-peer carsharing as well as pay-per-mile

pricing, have fueled much of the recent growth (13). One-way carsharing gives users more

flexibility by allowing cars to be dropped off at a different location from where they were picked

up. Rather than being assigned to specified pods, these vehicles often circulate freely through

cities. Austin-based car2go, owned by Daimler AG, is the market leader in this category, now

offering membership in nine metropolitan areas (10). Another relatively new service, peer-to-

peer carsharing, allows users to rent cars owned by individuals living nearby.

The growth in the traditional neighborhood model for carsharing, however, appears to

have plateaued – or even reversed itself– over the past year. According to Susan Shaheen and

Adam Cohen of the University of California-Berkeley, the number of available carsharing

vehicles (not including peer-to-peer vehicles) rose from 16,811 in mid-2013 to 19,115 in mid-

2014, but then dropped to 16,754 by early 2015 (13).

Some of this 11.7% decline may be due to seasonal issues related to the timing of the

5

counts. Analysis by these researchers, however, also points to modest declines in membership

between 2014 and 2015. One reason could be that carsharing apparently faces rising competition,

including that from ridesourcing companies such as Lyft and Uber (discussed in greater detail

below), as well as innovations by traditional car-rental services, which have begun to allow for

more short-term rentals.

Despite such mounting competition, relatively little research exist on the prices and

demand elasticity of carsharing. Numerous studies survey consumers about the importance of

cost in their decision to carshare in specific locations (3, 4 & 8). Shaheen, Cohen and Roberts in

2006 evaluated membership fees of 26 carsharing organizations, and assess the implications on

the sector’s ability to grow. Nonetheless, little available research exists about the actual prices

charged across the sector. Among the few that explore price changes is Jiangping Zhou, who

evaluates a campus program and estimates that the elasticity of demand created by “free hour”

promotions is -0.57, suggesting that each 10% increase in cost will diminish use by 5.7% (16).

The author notes that the overall elasticity of demand has likely risen since this data was

collected in 2011.

III. METHDOLOGY

To measure the prices and tax burden facing carsharing, the lowest priced vehicle available

and taxes and fees added by government agencies were recorded at locations throughout the

United States. Reservations were made for a compact vehicle, such as a Honda Civic, which are

common across fleets. The sample consists of price measurements at a representative pod in 80

cities falling into two categories: i) the 40 most populous U.S. cities in which neighborhood

carsharing is available through car2go, Maven, and/or Zipcar, all of which have populations of

300,000 or more, and ii) a stratified sample of Zipcar pods in 40 smaller cities ranging in

population from 4,000 – 299,999. This sample of cities is listed in the appendix.

In each of these 80 cities, a neighborhood pod in relatively close proximity to the city’s

downtown maintained by Zipcar was selected. A criteria for pod selection is having a heavy

orientation towards a residential population. Pods located in central business districts were

excluded to avoid selecting higher-cost services that are not representative of the sector as a

whole. By excluding airport and downtown locations, the analysis avoids the special taxes and

surcharges applied to transactions that do not apply in neighborhood settings. In these cities,

Zipcar prices and taxes in 32 pods can be compared to the author’s identically-collected sample

from spring 2011 (1). The sample of Zipcar pods taken in 2011 was smaller partially because the

company had a smaller network at that time.

Data collection for price and tax information took place from April through June 2016.

Membership was required for each of the three companies to obtain the prevailing prices. In each

location sampled, reservations of three lengths—one hour, five hours, and 24 hours—were

sampled for the use of a car on Wednesday mid-afternoons (2 p.m.) about a week in advance.

Altogether, when both 2011 and 2016 data are combined, the dataset consisted of 280 base prices

and 758 tax amounts (generally, several different taxes are applied per reservation).

6

As discussed in Section II, the lowest price of car2go and Maven vehicles in close

proximity to the Zipcar pods was also measured. Enterprise Carshare was not included in the

comparisons since it does not allow members to readily view prices outside their home city.

IV. CHANGING BASE PRICES, 2011-2016

In the 32 cities with identical pod locations in which the hourly prices for Zipcar were

available in both 2011 and 2016, the results show a noteworthy drop in prices. The base price of

a one-hour Zipcar carshare (weighted on the basis of city population and not including taxes) fell

from $9.39 to $8.92, a decline of 5.0% (Table 2). (This difference is significant at a .01 level of

significance). The real decline, i.e., the drop after prices are adjusted in accordance with the

consumer price index, is even greater at 11.2%.

Of the 32 pods with identical locations, nearly twice as many experienced nominal price

declines as price increases. Twenty (63%) experienced declines, one did not change (3%), and 11

(34%) had experienced price escalation. In real terms, even more pods experienced declines.

The pattern is exemplified in the shifting interquartile range, which went from $8.00 - $9.50 to

$7.50 - $9.38.

The percentage decline is almost as high when the 2011 prices are adjusted for the

changing cost of car ownership and driving, which is measured annually by the Internal Revenue

Service (IRS). The agency estimates that costs rose from $0.51/mile in early 2011 to $0.54/mile

in 2016. When adjusted in this way, the cost of carsharing is found to have dropped 10.3%

relative to the cost of private automobile use. The average gasoline price nationwide,

meanwhile, fell from $3.51 in June 2011 to $2.25 in the same month in 2016 (U.S. Energy

Information Administration, 2016). Carsharers, however, tend to use more fuel-efficient vehicles

and are more likely to use electric cars than the typical car owner, which reduces the sensitivity

of hourly fees to fuel costs.

In both years, rates in small cities were slightly lower than those in large cities. Larger

cities also experienced a much greater drop in prices than smaller ones. Among cities with

populations of 500,000 or more, prices fell from $9.53 to $8.97, or 5.8%. However, those in

smaller cities declined from $8.61 to $8.60, just 0.2%.

When the analysis is expanded to consider all 80 Zipcar pods sampled in 2016, the

average price was $8.57/hour. The highest prices were found in Charlotte, NC ($12.25) and

Berkeley and Oakland, CA ($10.75), and among the lowest were seen in Detroit ($5.50) and

Evanston and Oak Park, IL ($7). Table 2 also shows the average prices for five-hour ($42.18)

and 24-hour reservations ($72.39).

7

TABLE 2

Changes in Average Price of One-Hour Carsharing, 2011 – 2016

Zipcar Pods, Weighted by City Population

Changes in 32 Pods with Data for 2011 and 2016

2011 2016 % Change

Nominal Rate $9.39 $8.92 -5.7%

Real Rate (2011 prices adjusted to

2016 dollars)

$10.05 $8.92 -11.2%

Cost adjusted (2011 prices

adjusted to 2016 driving cost)

$9.94 8.92 -10.3%

> 500,000 population

$9.53 $8.97 -5.8%

< 500,000 population

$8.61 $8.60 -.2%

Changes in all 80 Cities Sampled in 2016

1 hr. average

5 hr. average

24 hr. average

All pods sampled

$8.57

$42.18

$72.39

The methodology does not allow for a determination of whether the price changes were

driven by shifts in supply or demand, or the sum combination of these factors. It appears likely,

however, that on the supply side, rising economies of scale are fostering greater efficiencies.

The estimated size of the U.S. carsharing fleet has more than doubled between 2011 and 2015

(13). On the demand side, rising competition – both among carsharing organizations and

between modes – also appears be to an explanatory factor.

Related to this latter factor, technological innovation has greatly expanded consumer

choice. Whereas five years ago, consumers often had a choice of owning a car, taking public

transit, walking or biking, or carsharing can often now also: i) summon a ridersourcing service,

such as Lyft or Uber, which provide the benefit of curbside pickup with a simple smartphone

click; ii) hail a carpooling service, such as UberPool or LyftLine, which now operate in many

cities and considerably reduce the cost of a typical ride in exchange for serving other passengers

during a trip; and/or iii) rely on specialized minibus or shared-ride van services that are now

prolific in many corridors. Bikesharing programs are also on the rise and now boast hundreds of

“pods” in major cities, providing new competition, especially on short-distance trips. These ever-

improving alternatives appear to be making the demand for carsharing more price-elastic.

8

Differences in Price between Operators

Measuring price differences between operators is relatively complex due to the various

approaches to pricing that each operator employs. Whereas car2go emphasizes per-minute

prices, Maven and Zipcar (as well as other national players, such as Enterprise) emphasize

hourly reservations, although several, including Zipcar, are experimenting with a pay-per-minute

approach (15). Another difference is that Zipcar tends to operate from pods with a variety of

cars while car2go and Maven tend to have cars parked individually, including in on-street

parking spots throughout the city. To assess the approximate difference in price, the analysis

considers the lowest observed rate of the car2go and Maven vehicles within a two mile radius of

the Zipcar pod evaluated. If no cars are available in that zone, the price of car at the 25%

percentile in terms of price for that city was selected.

Maven and Zipcar were found to have similar pricing levels. In the four markets served

by Maven (see Table 1), the average weighted price was $8.79/hour compared to Zipcar’s

$8.43/hour (Figure 1). Both companies appear quite aggressive in adjusting prices on the basis of

local conditions. For example, both had sharply discounted prices in Washington, DC, charging

$6/hour or less—about half that of other cities.

The results also show that car2go has more uniformity in pricing across its locations. The

operator generally charged $0.41/minute or $14.99 per hour, whichever is less (car2go serves the

nine cities listed on Table 1). All prices were set at this standard level except in two cities, where

prices were lowered to $0.19 and $0.31/minute, or $10 per hour. Overall, the average car2go

hourly price was $13.58 compared to Zipcar’s $9.13.

These observed differences, however, are somewhat deceptive, as car2go’s per-minute

charges (which averaged $0.385) allow consumers to more precisely match their payments with

the time they spend driving. As shown in Figure 1, car2go consumers who drive 20 minutes or

70 minutes would enjoy savings of $1.44 and $.83, respectively, over Zipcar. Some of the above

differences, of course, may be explained by factors related to size of cars and other vehicle

choices made available. When considered broadly, however, the results suggest that the least-

cost option will depend heavily on the characteristics of the trip.

9

FIGURE 1

Average Base Prices in Overlapping Markets

by length of trip; not including taxes

None of the companies offered significant discounts for five-hour reservations. For 24-

hour reservations, car2go tended to offer steeper discounts, charging the equivalent of five hours

for a full day, while Maven charged eight hours and Zipcar averaged 7.7. The data also shows a

modest correlation (.135) between the hourly price of a one-hour Zipcar reservation and a city’s

population. A moderately negative correlation (-.186) exists between cost of living and price.

The correlation between a city’s population density and price, however, is weak (0.052).

Multiple Regression Model

A multiple regression model helps illustrate how price differences vary by company and

the city’s population. The dependent variable is the hourly price (which, as previously noted, is

for the least expensive vehicle available), while the independent variables are the city’s

population and the company offering the car. Several other variables, such as population density

and cost of living were also tested but found to be less effective control variables.

This regression equation below shows an intercept of $8.11 per hour and a pricing rise by

$0.22 for each million residents of a city. Both coefficients are statistically significant at a .05

level, although population only marginally so. The latter coefficient suggests that a city with two

million residents will have hourly prices about a half-dollar higher than in small communities.

$8.79 $8.43

$13.58

$9.13

$7.69

$9.13

$17.43 $18.26

$0

$2

$4

$6

$8

$10

$12

$14

$16

$18

$20

Maven Zipcar car2go Zipcar car2go Zipcar car2go Zipcar

-70-minute trips-

----------- One-hour trips -----------

-20-minute trips-

10

Controlling for population, Maven’s prices are $1.11 higher than Zipcar’s (which is the reference

category), although this coefficient is not statistically significantly different. By contrast,

car2go’s hourly prices are $3.54 higher than Zipcar’s, a statistically significant difference.

Hourly Price = $8.11 + . 23 (POP) + $1.11 (Maven) + $3.54 (car2go) + e

(.000) (.046) (.13) (.000)

R2 = .438

Adjusted R2 = .419

Obs. = 90

With p values in Parentheses

Variables

Hourly Price Lowest hourly price, weekday at 2 p.m.

POP City population in millions (2010 census)

Maven Maven price observation (dummy))

car2go car2go observation (dummy)

Zipcar is the reference category

These results, together with an R2 value of 0.438, suggest that this simple model is a reasonably

robust predictor of price differences between firms and cities of different sizes.

Caution should be exercised in interpreting the Maven coefficient since it involves only

four observations, and the coefficient for car2go is partially a function of its differing fee

structure. Nonetheless, taken as a whole, these results illustrate the dynamic relationship

between a city’s size and the company offering the car.

V. TAX CHANGES

The second objective of the study is to measure the level of taxation, and trends in

taxation, since the 2011 sample was drawn. Previous research shows that carsharing’s status as a

relatively heavy taxation sector is largely the result of the once-prevalent notion that the

incidence of taxes on car rentals will fall almost entirely on out-of-towners, including tourists,

business travelers, and conference goers (1). Many taxes collected on car rentals indeed target

those living outside the jurisdiction imposing the tax. As carsharing services expand, however,

these taxes are being increasingly felt locally.

The results show that transaction-based fees (i.e., flat fees regardless of the base price or

duration of the reservation) are quite prevalent for carsharers. Transaction fees of $2.00 - $2.50

are most common. In Columbus, OH, users pay a $4 city vehicle lessor tax in addition to other

taxes (Figure 2). Pittsburgh, PA users face two flat $2 fees. New Jersey’s flat $5 fee is the

highest in the 80 pods sampled. Boston, MA has a $10 fee, but carsharing members pay this

only on their first reservation annually.

11

12

TABLE 3

Tax Rates on Carsharing in 12 Largest U.S. Cities Ranked by Population; Highest quartile in bold

1 HR.

TAX

5 HR.

TAX

24 HR.

TAX

GENERAL

SALES TAX

1 New York, NY 19.9% 19.9% 19.9% 8.75%

2 Los Angeles, CA 8.9% 9.0% 9.0% 9.0%

3 Chicago, IL 21.2% 22.7% 24.3% 10.25%

4 Houston, TX 15.1% 15.0% 15.0% 8.25%

5 Philadelphia, PA 33.8% 24.3% 14.7% 8.0%

6 Phoenix, AZ 48.9% 34.1% 19.2% 8.6%

7 San Antonio, TX 15.1% 15.0% 15.0% 8.25%

8 San Diego, CA 8.0% 8.0% 8.0% 8.0%

9 Dallas, TX 10.0% 10.0% 10.0% 8.25%

10 San Jose, CA 8.8% 8.8% 8.8% 8.75%

11 Austin, TX 15.1% 15.4% 15.0% 8.25%

12 Indianapolis, IN 17.1% 17.0% 17.0% 7.0%

The average tax across the 32 cities rose from 15.5% in 2011 to 17.0% in 2016. This 1.5

percentage point increase indicates that rising taxes have offset approximately 30% of the

reduction in prices over this period. When looking at the entire sample of 80 pods for 2016,

including those for which data for 2011 is not available, the average tax rate was found to be

even higher, 19.1%, partially due to the inclusion of several highly taxed New Jersey pods.

Eight of the 12 largest cities impose taxes of 15% or more on all types of reservations

(one, five and 24 hours). In two of the three largest cities, Chicago and New York, rates hover

around 20%, while in Philadelphia, PA and Phoenix, AZ they exceed 33% in some scenarios

(Table 3). Tax rates are more favorable in Dallas, Los Angeles, San Diego, and San Jose, where

rates are near or only moderately above sales tax rates.

Rates tend to drop as the duration of the reservations increase due to the diminishing

impact of flat-fee taxes on more expensive purchases. In most cities, rates are well above those

for sales tax. In Philadelphia, taxes are more than three times the sales tax rate for reservations of

five hours or less. In Phoenix, one-hour carsharers pay almost half as much in taxes as they do

for the car itself—and about six times the rate of the sales tax. In five of these 12 cities, rates of

taxation can be at least 50% higher than local sales taxes.

Weighted Average Tax Rates in 40 Largest Cities

The differing tax burdens facing consumers in the 40 largest cities can be shown by

considering the weighted average tax rate (Table 4). The cities are assigned into three categories

based on population and ranked among the entire sample of cities as well as those in their size

13

TABLE 4

RANKING OF CARSHARING TAXES IN 40 LARGEST U.S CITIES

Based on mix of 45% one-hour, 45% five-hour, and 10% 24-hour reservations

Tax Rank Tax Rank Weighted Average Sales

in Category 40 Cities City Tax Rate Tax Rate

Cities with populations of more than 1,000,000

1 2 Phoenix, AZ 34.0% 8.6%

2 10 Philadelphia, PA 24.1% 8.0%

3 12 Chicago, IL 21.4% 10.3%

4 14 New York, NY 19.9% 8.9%

5 19 Houston, TX 15.0% 8.3%

6 20 San Antonio, TX 15.0% 8.3%

7 29 Dallas, TX 10.0% 8.3%

8 33 Los Angeles, IL 9.0% 9.0%

9 36 San Diego, CA 8.0% 8.0%

Cities with populations of 500,000 – 999,000

1 3 Columbus, OH 32.3% 7.5%

2 5 Tucson, AZ 30.8% 8.6%

3 6 Jacksonville, FL 26.9% 7.0%

4 7 Washington D.C. 26.5% 5.8%

5 8 Denver, CO 25.5% 7.7%

6 9 Albuquerque, NM 24.5% 7.2%

7 13 Las Vegas, NV 20.1% 8.2%

8 16 Seattle, WA 17.3% 9.6%

9 17 Austin, TX 15.2% 8.3%

10 18 Milwaukee, WI 15.3% 5.6%

11 22 Charlotte, NC 14.5% 7.3%

12 23 Memphis, TN 14.3% 9.3%

13 24 Oklahoma City, OK 14.3% 8.4%

14 25 Indianapolis, IN 17.0% 7.0%

15 26 Nashville, TN 13.4% 9.3%

16 28 Baltimore, MD 11.5% 6.0%

17 30 Fort Worth, TX 10.0% 8.3%

18 34 San Jose, CA 8.8% 8.8%

19 35 San Francisco 8.8% 8.8%

20 37 Detroit, MI 8.0% 6.0%

21 38 Boston, MA 6.3% 6.3%

22 39 Louisville, KY 6.0% 6.0%

23 40 Portland, OR 0.0% 0.0%

Cities with populations of more than 300,000 – 499,000

1 1 Fresno, CA 39.5% 8.8%

2 4 Pittsburgh, PA 30.7% 7.0%

3 11 Minneapolis, MN 22.1% 7.8%

4 15 Omaha, NE 16.8% 7.0%

5 21 Miami, FL 15.0% 7.0%

6 27 Sacramento, CA 14.7% 8.8%

7 31 Oakland, CA 9.5% 9.5%

8 32 Atlanta, GA 9.3% 8.0%

14

category. The weighted averages are based on a scenario in which 90% of use is divided equally

between one and five-hour reservations, and the remaining 10% are 24 hours in duration.

Thirty of the 40 cities have retail taxes of 10% or more. Among the largest cities,

Chicago, Philadelphia and Phoenix have the highest average taxes, while taxes are lowest in San

Diego. Columbus, OH, and Fresno, CA also are among the highest due to lump-sum taxes. In

only four of the 40 cities – Louisville, KY and Portland, OR, as well as Los Angeles and San

Diego, CA – are tax rates equal to or less than sales taxes. Portland is notable for exempting

carsharing from taxes.

Cities Increasing Tax Rates, 2011 and 2016

The cities with appreciable increases for one-hour reservations include Chicago (+9.2

pts); Omaha, NE (+8.9 pts); Minneapolis (+8.1 pts) and Milwaukee (+3.8). Among large

markets, Chicago stands out for its sizeable increase. Prior to 2015, carsharing organizations that

could meet certain criteria were exempted from the city’s lease tax. Following the acquisition of

the nonprofit I-GO by the for-profit Enterprise, this waiver was lifted, pushing tax rates on most

transactions from about 12% to 21%.

The 17.0% average rate of taxation on carsharing is slightly higher than the 16.7% tax

rate on two-day neighborhood car rentals in the same 32 cities. (The latter estimate was made by

computing the average tax rate for a two-day reservation of an Enterprise compact car from a

non-airport location in the same 32 cities.) Both of these rates, in turn, are well above those for

other intercity transportation services, such as airline tickets. 16.0%, including security fees, as

estimated by the Business Travel Coalition (2).where is reference) and intercity bus and train

fares (which are not taxed). Even hotel taxes, as estimated by Hazinski (5), are lower, averaging

about 13.5% nationwide.

This suggests that carsharing finds itself in a far less favorable position with respect to

taxes than nearly all other sharing economy sectors. It competes with other urban transportation

services, including ridesourcing (Lyft/Uber) and bikesharing, which are generally not subject to

retail taxes. Although there are exceptions in some cities, and data is not available on the mean

rate of taxation on ridesourcing and taxicabs, the median rate of taxation is zero for both service

VI. CONCLUSION

The significant fall in the price of carsharing suggests that consumers are reaping benefits

from the sector’s evolution and expansion. In addition to being more environmentally sustainable

than car ownership, carsharing is increasingly becoming more financially sustainable for the

many lower-income consumers its providers serve. Maven and Zipcar have similar pricing

philosophies that emphasize low hourly rates, while car2go employs a per-minute rate with a

somewhat higher hourly price.

The decline in rates follows a pattern common among infant industries that start with an

experimental business model and gradually benefit from greater economies of scale, heightened

product awareness, and growing managerial sophistication. The marked decline in one-hour

15

rates, together with rising taxes, however, suggests that companies are likely facing significant

pressure to contain costs for the lowest-priced vehicles. Taxes increases have offset almost a

third of the price decline, which may negatively affect operating margins.

Whether taxation is contributing to the recent slowdown in the growth of carsharing in

major U.S. cities has not been systematically evaluated. More research on the effects of taxation

on consumer demand is clearly needed, particularly with respect to transaction fees imposed by

many locales. The issue of differential rates of taxation between carsharing and other shared-

economy sectors is coming into sharper focus as the number of mobility options available to

urban dwellers grows. It is incumbent on carsharing organizations to raise awareness of the

potentially deleterious effects of such fees on their ability to grow.

16

Appendix Cities with Neighborhood Carsharing in Sample

Data collected for all 80 cities for Zipcar. C denotes car2go data also collected; M denotes Maven collected

17

VII. REFERENCES

1. Bieszczat, A & J. Schwieterman, (2013). "Carsharing: A Review of the Public Benefits and

Tax Burden of an Expanding Transportation Sector," Journal of the Transportation Research

Board.

2. Business Travel Coalition, (2015). “Are U.S. Airlines Overtaxed,” December. Accessed at

http://www.businesstravelcoalition.com/documents/are-us-airlines-overtaxed.html

3. Costain, C., C. Ardron, N.K. Habib, (2012). “Synopsis of users’ behavior of a carsharing

program: A case study in Toronto,” Transportation Research Part A: Policy and Practice, 46

(3): 421–434.

4. Econsult Corporation, (2010). “The Economic and Environmental Impact of PhillyCarShare in

the Philadelphia Region.” , Econsult Corporation.

5. Hazinski, T.,Moon, A. & Rees,M. (September, 2015). HVS Lodging Tax Report, HVS

Convention, Sports, and Entertainment.

6. Lindzon, Jared, (February, 25 2015). “Why auto makers encourage car sharing despite trend away from ownership,” The Globe and Mail.

7. Litman, Todd, ( September 17, 2015). Evaluating Carsharing Benefits, Victoria Transport Policy

Institute.

8. Lane, Clayton, (2005). “PhillyCarShare: First-Year Social and Mobility Impacts of Carsharing in

Philadelphia, Pennsylvania” Transportation Research Record: Journal of the Transportation

Research Board, no. 1927: 158-166.

9. Martin, Elliot W. and Susan A. Shaheen. (2011) "Greenhouse Gas Emission Impacts of

Carsharing in North America" IEEE Transactions on Intelligent Transportation Systems 12, No.

4: 1074-1086.

10. Martin, Elliot and Susan Shaheen,( 2016). “Impacts of car2go on Vehicle Ownership, Modal

Shift, Vehicle Miles Traveled, and Greenhouse Gas Emissions: An Analysis of Five North

American Cities”, Transportation Sustainability Research Center, University of California –

Berkeley.

11. Millard-Ball, A., Murray, G., Ter Schure,J., Fox , C., & Burkhardt, J.( 2005). Carsharing: Where

and How It Succeeds. Transit Cooperative Research Program (TCRP) Report 108, Washington,

DC: Transportation Research Board. Accessed August 27, 2010.

12. Scott, S., Brook, D., & Perussi, M.( 2003). “Impacts of Carsharing on Walking Behavior.” Paper

presented at the Walk 21 conference Portland, Oregon, May 1-3. Accessed December 18, 2010.

13. Shaheen, Susan A., & Adam Cohen, (2015). “Innovative Mobility Carsharing Outlook: Market

Overveiw, Analysis and Trends”, Transportation Sustainability Research Center, University of

Calironia Berkely.

14. Shared-Use Mobility Center/American Public Transit Association,. (2016). Shared Mobility and

the Transformation of Public Transit (TCRP J-11/Task 21), 2016. Submitted by the Shared-Use

Mobility Center to APTA.

15. Steele, Brian, ,( June 28, 2016). “Zipcar Tests Pay-Per Miles Pricing for Short Trips”, Engaget

available at https://www.engadget.com/2016/06/28/zipcar-tests-pay-per-mile/

16. Zhou, Jiangping.(2014). Carsharing on University Campus: Subsidies, Commuter benefits, and

their impacts on Carsharing. Transportation Research Part D 32 316–319

17. Zipcar (2013). "zipcar reports fourth quarter and full year 2012 results", Press release.

18. zipcar.com. Official Zipcar website, available at www.zipcar.com