Transportation Deployment Casebook/2014/US Carsharing Vehicles

Carsharing in the U.S.
Carsharing is a type of mobility service that allows members to easily rent or use vehicles for short periods of time. Carsharing services grew out of a desire to access cheaper car usage or to avoid car ownership for financial, environmental, or lifestyle reasons. Carsharing services have the potential to reduce auto ownership in urban settings and better utilize available parking and vehicles.

Definition and Attributes
Most carsharing services share certain features:

This report will focus on two general types of carsharing services: point-to-point and round trip, station-based. Point-to-point services, such as Car2Go, allow for one-way travel. Cars can be picked up from and dropped in different locations. In the case of Car2Go, trips are ended when a car is successfully parked in any non-private parking spot with a greater than 2 hour time limit that is within the Car2Go’s parking boundaries. Round trip, station-based services have predetermined parking spaces; cars can be checked out from and returned to the same “dock.” Car companies that have used this approach include ZipCar and locally-based HourCar. Round trip carsharing services have been around for longer than point-to-point services. Users generally reserve a car in a particular location for a particular amount of time. Fares are based on the total time reserved, and returning a car late can have expensive consequences. Cars are managed, owned and maintained by a central company. There is a third type of carsharing service that is currently being implemented by Zipcar in Boston, MA. Called point-to-point station-based, it is a service that can be used for one-way trips but cars must be returned to a number of dedicated parking spots, often with infrastructure for recharging electric vehicles. While this is an easier system to manage operationally than point-to-point “free floating” services like car2go, these services offer a lower degree of flexibility for users.
 * Require one-time membership and screening wherein a potential driver has their identity and driving record verified.
 * Many services are keyless. Cars are accessed via either a fob or membership card.
 * The user is the driver, unlike a taxi where a driver is provided as part of the service.
 * User fees are based on the time the vehicle was used or distance traveled.
 * Cars are accessible 24 hours a day.

Invention
The first record of a formal carsharing program began in Zurich, Switzerland in 1948. Called “Sefage,” it was started by a group of individuals who could not afford to purchase a car, so instead agreed to share cars. Other “public car” experiments were attempted in France and Amsterdam in the early 1970’s, but ultimately failed.

The first carsharing companies began in European countries in the late 1980’s and early 1990’s, but many of these initial companies either failed in the long-term or were acquired by more professional emerging companies. Initial programs were funded by government grants and operated on a neighborhood scale. Expanding these smaller companies into larger markets proved challenging. Many companies misestimated the number of vehicles necessary, overemphasized infant technologies, and failed at effectively marketing the services outside of the original jurisdiction.

In the U.S., experiments in carsharing ventures began in the 1980’s. Purdue University introduced Mobility Enterprise while San Francisco tried a program named the Short-Term Auto Rental (STAR) initiative. Purdue’s experimental program failed to produce long-standing results, and both Mobility Enterprise and STAR were abandoned in 1985 and 1986, respectively.

Early Market Development
It wasn’t until 1994 that carsharing companies in the U.S. emerged. These were structured as either non-profit (including cooperatives) or private companies. The first of these, Dancing Rabbit Vehicle Cooperative began in 1997 in Rutledge, MO. It opened on a cooperative, non-profit business structure. CarSharing Portland, a carsharing pilot project started by the Oregon Department of Environmental Quality and the U.S. Environmental Protection Agency, was the next U.S. company to begin operations (March 1998). Though initiated through government funding, it operated as a for-profit business until being acquired by Flexcar in 2001. Flexcar (January 2000) is a carsharing company based out of Seattle which was launched in part due to strong political backing, but operates as a private commercial venture.

Two other private companies were started in 2000: Boulder CarShare in Boulder, Colorado and CarSharing Traverse in Traverse City, Michigan.

ZipCar began in June 2000, funded largely through venture capital funds. Originally based out of Boston, ZipCar was modeled off of similar carsharing services in Europe. ZipCar had the advantage of being created and deployed following the rise and fall of similar European companies, and thus avoided many of the deployment mistakes of earlier companies. As of 2005, ZipCar had 800 members and 42 vehicles.

From 1994 to 2005, 40 carsharing programs were deployed in the U.S., the majority of which were located in urban areas. The number of companies entering the carsharing market boomed between 1999 and 2001 (9 programs in 2001 alone). The entry of new companies has since stabilized. Barriers to new companies include the potential for larger companies to expand into new markets (cities). Larger companies benefit from economies of scale and first-to-market advantages.

Membership growth rates slowed between July 2004 and July 2005. Memberships had doubled every year leading up to this point since 2001. Nationwide annual vehicle growth for carsharing companies maxed out at 30% in 2005, with the largest operators responsible for the majority of membership and fleet growth. These larger companies also had the highest member to vehicle ratio (66:1) as compared to the average member-vehicle ratio of smaller companies (20:1). Higher membership-vehicle ratios increase per vehicle use, profitability per vehicle, and encourages further investment. The nationwide average membership-vehicle ratio in 2005 was 64:1. That this average is so close to the average of the large companies exemplifies that these companies hold the majority of the market.

High membership numbers may also be a result of limited membership requirements. Many U.S. companies have one-time fees for signing up while some have monthly or annual fee requirements.

Niche Market
The carsharing market has been described relying on one niche market, with potential to break into a number of other niches. Users tend to be:
 * Users of non-car forms of transportation in urban environments
 * Residents of urban neighborhoods
 * Living in carless or single-car households
 * Middle or middle-to-upper income
 * Not married or living in a childless-couple household
 * Young adults, primarily between the ages of 25 and 45
 * Predominantly male
 * Well-educated

Researchers estimate that development of carsharing services for low-income people who do not drive much but would like access to a car sometimes, elderly people who do not want the responsibility or financial burden of owning and maintaining their own vehicle, and commuters who value inexpensive parking at prime locations are market niches that are ripe for development. Among current carshare users, convenience and cost were the main reasons why people chose to be carshare members (56% of respondents). Populations in the previously mentioned niches are likely to be highly motivated by cost (low-income and elderly) and/or convenience (commuters looking for prime parking opportunities, elderly who are unable to walk long distances to reach shopping and commerce centers).

A survey conducted in 2005 polled carsharing organizations regarding current market trends and predicted future trends. The majority of U.S. companies responded that the primary market was residential-based and the targeted age was 21-25 year olds. The survey predicted that the most growth in memberships and use would continue to occur in residential neighborhoods. Some growth is expected in the business and college market (22% and 23%, respectively) and among low-income, commuter and older adult demographics.

The rise of insurance rates post 9/11 has caused some concern among U.S. carsharing organizations. Some companies have responded by lowering liability insurance, utilizing self-insurance of vehicle damage, or by forming carsharing affinity groups, a type of insurance pool that offers reduced rates for members. High insurance costs presents challenges for companies seeking to break into the University student market, as insurance for younger drivers tends to be higher than more experienced drivers. Some colleges have dealt with this financial barrier by providing insurance through their liability policy in order to open up their campuses to carsharing ventures. Other universities have offered free or reduced costs parking on campus and have subsidized student memberships to entice carsharing companies.

Some public agencies have discussed the use of carsharing services to replace non-essential public vehicles used by staff. Recently, New York replaced 50 vehicles with contract for 25 ZipCars. It’s estimated that this will save the city $500,000 over four years. Chicago implemented a carsharing contract with ZipCar for its staff in 2011, and has saved over $3 million dollars according to Kevin Campbell, Chicago’s manager of fleet services. San Francisco is considering shifting to a similar program, largely for economic reasons.

ZipCar announced plans recently to expand service to airports to target business customers that may be in town for the short-term. Recently acquired by Avis Budget (2013), the rental car company has taken advantage of shifts in rental and carsharing demand that occurs throughout the week or seasonal swings. Avis transferred 1,000 vehicles to ZipCar to meet summer demand. Likewise, Avis Budget has higher demand during the workweek, while ZipCar’s demand is concentrated more highly on the weekends. It is estimated that Avis Budget Group will gain over $20 million in cost savings through better fleet utilization.

As the lines continue to blur between rental cars and carsharing vehicles, rental car operations are expected to integrate some of the best practices of successful carsharing operatives. Researchers at the Transportation Sustainability Research Center at the University of California - Berkeley predict that rental companies will be more likely to do pre-screening or one-time membership processes like those of carsharing companies. Rental processes are also expected to take on some of the attributes of carsharing processes, such as automated or unattended car rental services.

Technology
Technology has played a major role in both company rollout and long-term success. Using advanced technologies (e.g. automated reservations, integrated billing, and keyless vehicle-access systems) have allowed carsharing companies to retain a greater share of the market than smaller systems with fewer technology advantages. Available technology has allowed for improved customer experience, fast to instant billing, and lower operational costs as compared to companies that take reservations by phone. In 2005, 55% of respondents either offered or were considering offering instant reservation systems.

As the carsharing market develops, technology utilized by carsharing companies is expected to spread into parent rental car companies. Telematic technology is expected to be integrated into rental car fleets so that rental cars can be rented by the day or by the hour and have unattended virtual or keycard access, allowing an entire fleet to be more flexible in response to market demand. Cloud-based navigational and vehicle locator devices offer improved capabilities for vehicle management and operations.

Policy
Privileged access to on-street parking spaces, often in desirable, dense residential or commercial environments has been described as a critical vulnerability of carsharing services. Public-sector control of municipal, on-street parking spaces carries several implications for carsharing companies. Public-sector entities often implement policies and decisions at slower paces than the private sector, are under pressure from a range of constituents and groups to produce a large range of outcomes, and are highly susceptible to changes in policy direction and party control. Agencies have formed to manage the complex and necessary relationships between municipalities and carsharing services. In the U.S., the Shared Use Mobility Center was introduced in June 2014. Another of these groups, the European Automobile Manufacturers Association (ACEA), recommends developing these public-private relationships on an industry level, rather than each carsharing company approaching each highly-fragmented municipality.

Carsharing companies in many states are still subject to be taxed according to rental car tax policies. Rental car tax policy reasoning, which is seen as a politically acceptable way to increase revenues coming from unrepresented visitors, is not fundamentally applicable to the early users of carsharing services. Most are local residents, which are outside of the original purview of car rental tax policy.

Current Market
As of August 2014, there were 1,228,573 members and 24 car-sharing operators in the U.S. The total U.S. carsharing fleet was 17,179 vehicles, which makes a 71:1 membership to vehicle ratio, significantly higher than it was in 2005. Researchers predict that market share has the potential to occupy 10% of the market for individuals over the age of 21 in North America.

Quantitative
Data was taken from a report titled “Innovative Mobility Carsharing Outlook” from the Transportation Sustainability Research Center at the University of California, Berkeley. Birth stage data was found in “Carsharing in the United States: Examining Market Potential" an earlier paper written by the same author . The oldest data is from 1998 and has been voluntarily collected from carsharing companies in July of each year.  Data includes one-way and round-trip carsharing.

The innovation of a new technology can be described with an S-curve. There are four major parts of every S-curve and corresponding innovation: birth, growth, maturity, and decline. The birth phase is categorized by slow growth as new systems find their place in the market. After reaching a certain place in it’s lifecycle, people become much more open to its presence in the market, and begin to use the service at an exponential rate. Eventually, this growth rate is slowed due to any combination of factors: physical constraints, saturation of market, emergence of new technology, etc.). As new technology continues to innovate, the old technologies decline and the new technology takes the place of the old in the market.

Because formal carsharing is a relatively new phenomenon, the data has not yet reached a full birth, growth, maturity and decline cycle. The number of vehicles in the U.S. that are owned by carsharing companies is just beginning to grow rapidly. Using this data in a three-parameter logistic function, a projected life cycle was established. S(t) = K/[1+exp(-b(t-t0)] where:
 * S(t) is the status measure, (i.e. vehicles)
 * t is time (i.e. years)
 * t0 is the inflection time (the year in which 1/2 K is achieved),
 * K is saturation status level, or the year that the system is overly mature
 * b is a coefficient that describes the steepness of the curve

K was estimated using researcher predictions that carsharing could capture 10% of the U.S. urban driving market between the ages of 21 and 45. U.S. Census data was used to find the number of valid drivers licenses in the U.S. (2009) for that age range. According to Census data, 80.7% of the population lives in urban areas, so eligible driver’s license numbers were multiplied by this urban factor. Lastly, since a membership to vehicle ratio of 66:1 was considered profitable by early companies, it was assumed that this rate or better would be sustained into the future. The number of drivers was divided by the number of memberships in the ratio to find the number of vehicles needed for that many drivers. K was estimated to be 113,484 vehicles as 100% saturation. ½ K, or the inflection point, was found to occur sometime in 2014. R-square had a value of 0.9098, which though not a perfect curve, fit the projected researcher estimates of market maturity values. Vehicles are predicted to reach K values in 2049, which fits with rule of thumb of lifecycle analysis: 50-60 years is a typical S-curve for transportation innovation.

The model underestimates initial years, and overestimates later years as compared to actual vehicles, which is expected from an Ordinary Least Squares Regression. Individual data point accuracy is second to best fit overall in priorities of calculation with this type of regression.