Feasibility and Consumer Dynamics of a Shared Transportation System

FEASIBILITY AND CONSUMER DYNAMICS OF A SHARED TRANSPORTATION SYSTEM

Thelma Snuggs, Department of Consumer Sciences and Retailing, Purdue University

Richard Feinberg, Department of Consumer Sciences and Retailing, Purdue University

Jennifer Meoli, Department of Consumer Sciences and Retailing, Purdue University

F. T. Sparrow, Automotive Transportation Center, Purdue University

Robert Whiford, Automotive Transportation Center, Purdue University

Michael Doherty, Automotive Transportation Center, Purdue University

ABSTRACT -

A mobility enterprise is a new transportation concept aimed at increasing the productivity of the automobile through use of micro-mini automobile in conjunction with a shared fleet of fuller sized vehicles. The objective of this research were twofold. The first was to assess the extent to which a market for this shared transportation system exists. The second was to understand how potential uses and detractors define the advantages and disadvantages of the enterprise concept. To that end, two community (West Lafayette, IN and Columbia, MD) and one national survey was completed. The major findings were: 1) a substantial market for a shared transportation system exists, and 2) one can significantly predict individual membership from a select group of subjective perceptions. Demographic characteristics, transportation behavior and attitudes were not necessary or sufficient for predicting interest.

Individual dependence and use of the single occupant automobile has been the focus of much attention by the public policy sector and rightly so. Automobiles account for about 30 percent of the total U.S. consumption of oil. It has been estimated that 32 percent of all automobile trips are work commutes and 75 percent of these trips are in cars with only the driver (U.S. Department of Transportation, 1973). Yet, actually, only a few studies have targeted the private car for intervention as a means of energy conservation by focusing on the individual driver. The research that has been done has focused on (1) determining the effectiveness of transportation conservation programs and influence attempts aimed at increasing participation in energy saving activities such as carpooling, mass transit, reduced driving (Foxx & Hake, 1977; Geller, 1982; Hake & Foxx, 1978; Reichel & Geller, 1981) and 2) determining individual personality attributes (e.g. demographics and attitudes) that predispose one to modify transportation behavior. Problematically, these attempts have failed to identify a single class of discrimating demographic, attitudinal or personality characteristics (Bultena, 1976).

Moreover, on the whole, transportation conservation technologies (e.g., ridesharing, carpooling) have failed. Effective conservation rests not only on improved engineering and technology, but on modified individual behavior. it is the individual who must make the decision to purchase and use any technology. The application of physical interventions for all energy conservation requires understanding the human factor.

Since it is increasingly clear that alternatives such as mass transportation will not be able to attract significant numbers of auto users what is needed is an approach that focuses on better use of the private automobile. Recently, the Institute for Interdisciplinary Engineering Studies at Purdue University has acquired mini-micro class automobiles for experimentation on improved auto productivity. These cars are involved in an experiment called the Mobility Enterprise. Since these automobiles are not yet available in the United States and are smaller than the smallest American car presently available, the problem of introducing mini-micro vehicles is great. However, because of their low price (less than $3500) and high gas mileage (45-60 m.p.g.) it is anticipated that mini-micros will be introduced in the United States within the next 2-3 years.

The Enterprise concept is based upon better matching of trip requirements to requirements of vehicle attributes. Basically, for a monthly fee, somewhat less expensive than the average cost of owning and operating a car, participants in the Mobility Enterprise will receive continuous access to a mini/micro car, adequate for most driving purposes (such as commuting to work and other local trips), and access to a shared fleet of larger cars and small trucks available for special purposes such as long trips and sizable personal hauling (Feinberg, Snuggs & Bauer, 1923; Doherty & Sparrow, 1984

Three features of a Mobility Enterprise -- retained auto autonomy, easy access to an expanded fleet of vehicles and reduced cost are believed to be the keys to its success. These features are interrelated. An Enterprise member's minimum attribute vehicle (in the case of the Mobility Enterprise the mini/micro vehicle) provides the participant with the most economical means of accomplishing his/her most frequent trips. When a trip can be made using the mini/micro, the member knows he/she can travel without delay. When a member's minimum attribute vehicle is inappropriate for a desired trip, he/she must seek access to the appropriate special purpose vehicle in the shared fleet. This process may involve delays, particularly if the vehicle is garaged elsewhere. It may also involve some advance planning, paperwork, and out-of-pocket expenses, depending on the structure and design of the Mobility Enterprise. Additionally, the possibility exists that the desired vehicle may not be immediately available when needed because of a prior reservation. Such departures from guaranteed access and instant gratification of private car ownership are aspects of the Mobility Enterprise that must be offset by clear benefits. Such benefits appear to be possible, since the Enterprise can offer several improvements: (1) a wider range of vehicles available for temporary use by an individual or increased mobility, (2) a less complex set - of criteria for buying a car, (3) trip and ownership economies that can be translated into more discretionary income or increased mobility, and (4) a more efficient use of society's scarce and expensive resources. In this county the practice and idea of sharing has found acceptance for farm machinery, boats, vacation homes and condominiums. However, while vehicle sharing has attracted attention and interest in Europe, it has not done so in the United States.

While it is clear that both the mini-micro automobile and the Mobility Enterprise offers significant economies for the individual (e.g-., saving money) and for society (e.g., saving scarce resources - oil, steel, etc.) it is not clear that either will be any more successful than other systems designed for conservation -- e.g., public transit and car pooling.

Since the viability of the Mobility Enterprise will depend on its acceptance and utilization by individuals, it is important to understand the consumer's point of view. The Mobility Enterprise will be better understood and more effectively designed when there is data on how potential users, supporters and detractors define its advantages and disadvantages, its significant and modifiable attributes, and their feelings toward it.

In order to increase understanding of individuals who might be interested in a shared transportation system and the benefits, advantages and attractive features of such a system, the present study had two objectives. The first was to determine if there is a viable market for a shared transportation system in the United States. The second was to assess transportation behaviors, demographic characteristics, and attitudes and reactions of a random sample of residents in Columbia, Maryland to better predict, and understand consumer dynamics of adopting a shared transportation system in that specific community.

STUDY ONE

Market Feasibility

The acceptance or market feasibility of the mobility enterprise concept has been the subject of three distinct surveys. Two of the surveys were of a very localized natured (Lafayette, IN, and Columbia, MD) designed to primarily test the feasibility of such an operation at a specific site. A third -- a national survey was designed to gauge the general acceptance or feasibility of the concept to the driving-aged population of the U.S.

The Social Research Institute of Purdue University conducted the local survey of the Lafayette/ West Lafayette, Indiana area. The sample consisted of a stratified (based on socioeconomic status) random sample of 300 households in the area. Tippecanoe County is a designated SMSA. The survey instrument was administered by personal interviews of 30-45 minutes each.

The Institute for Public Policy and Public Administration of Purdue University conducted the local survey of Columbia Maryland. The sample consisted of a stratified random mailed survey of 1000 households in Columbia Maryland. The four strata were selected based on socioeconomic status.

Columbia Maryland was chosen for a local survey because it had been selected as a possible Community for the placement of a demonstration project of a Mobility Enterprise.

Finally, the national survey was conducted in May 1983 through J.D. Powers and Associates of Westlake Village, California as part of their Automotive Consumer Profile, a tri-annual mail survey. The actual survey was administered through National Family Opinion, Inc. of Toledo, Ohio. The sample size, approximately 5,000 individuals in 3,000 households, yields a representative cross-section of the U.S. population, 16 years or older. Thus, the sample is representative of the U.S. personal vehicle fleet.

Respondents on all 3 surveys completed an extensive questionnaire covering demographics, personal vehicle ownership, attitudes and opinions of current events and issues, and automobile buying intentions. In addition, all surveys consisted of a description of a mobility enterprise and subjects were asked about their willingness to join such an undertaking. The results of the three surveys clearly show that to differing degrees substantial markets exist for a shared transportation system such as the mobility enterprise: 8'/.' nationally, 16% in the Lafayette area, and 25% in Columbia, Maryland. These results are somewhat surprising when considering the novelty of the concept, but support the feasibility of shared transportation systems in the United States.

STUDY TWO

Characteristics of Interested and Non-Interested Respondents

Since there appears to be a receptive market for a shared transportation system the second objective of this research was to better understand and predict the consumer dynamics of adopting a shared transportation system. To accomplish this the first task was the identification of individuals who were interested in joining and those not interested in joining a Mobility Enterprise. For this purpose, the large stratified random sample of residents of Columbia, Maryland (N=1000) was used. Families were mailed a transportation questionnaire (available upon request from the authors) which included questions about transportation behavior, history of automobile ownership, attitudes, demographics, and reactions to the Mobility Enterprise. A total of 343 responses (34%) were complete enough to be included in the analyses. Interest in the mobility enterprise was assessed by presenting a verbal description of a ''generic" shared transportation system and asking respondents whether they were interested in joining such a system. On the basis of the answer to that question subjects were assigned to one of two groups. Eighty-five individuals who indicated they "definitely would" or "probably would" be interested in membership is such a plan, were assigned to the "interested in joining group." Two hundred thirty-nine were individuals who indicated that they ''probably would not'' or "definitely would not" be interested in such a plan were assigned to the "not interested in joining group."

Discriminant analysis was used to empirically determine the significant and meaningful differences between groups. Discriminant function analysis has come to be a popular and powerful tool for selecting the best set of discriminative variables (e.g. Huberty, 1975). In such an analysis, the researcher selects a collection of discriminating variables that measure characteristics on which groups are expected to differ. The discriminant analysis provides several tools to help interpret the data: First, the analysis selects variables in order of their importance of discriminating groups. Second, the analysis measures the success with which the discriminating variables actually discriminate the subject population of interest.

RESULTS

[A complete set of tables are available upon request from the authors.]

Demographic Description

The characteristics for the samples are presented in Table 1. Columbia, Maryland is a well educated, middle class community. The sample characteristics do no deviate significantly from the demographic characteristics of the overall population.

Testing of Group Differences

Eighty-five individuals (25~1,0 of the survey population were "interested" in participating in a shared transportation system and 239 (70%`) were "not interested." Group differences were assessed by a discriminant function analysis performed on all 159 variables. Differences between the two groups were further evaluated by F tests when appropriate.

Table 2 represents all variables determined by the discriminant analyses, to have the "best" ability to differentiate between those who are "interested" or "not interested" in a shared transportation system. The results demonstrate that individuals interested in a shared transportation system can be distinguished from those not interested with only 7 variables. The 7 variables with a standardized discriminant function coefficient are provided in the table to form a significant function with a Wilk's Lamda of .60, significant at less than the .001 level, and a canonical correlation of.64. The Wilk's Lamda is a measure of the original variable's discrimnant power before it is removed by the discriminant function. The cononical correlation is the measure of association between the discriminant function and the composition of groups.

TABLE 1

DEMOGRAPHIC ANALYSIS

Along with their analytic function, the discriminant analysis also serves a classification purpose. Multiplying an individual's scores on the discriminating variables by the associated coefficients lead to the ability to predict the likelihood of an individual's membership in any of the groups. Applying the raw scores used to derive the function 85% of all cases are correctly classified - 92%, of those who were interested in membership in the shared transportation system and 83% who are not. This differentiation is statistically significant, p.001.

The data from the Lafayette/West Lafayette survey was used to independently cross-validate the results from the discriminant analysis. Using the 7 variables found to discriminate most significantly in the Columbia survey to predict and classify the Lafayette/West Lafayette respondents, 44 correct classifications (92%) in the "interested" group and 220 correct classifications (91',",) in the "not interested" group was realized. These cross-validation results offer support for the validity, stability and generalzability of the results obtained in the Columbia discriminat analysis.

TABLE 2

DIFFERENCES BETWEEN INDIVIDUALS, "INTERESTED" AND "NOT INTERESTED" IN THE SHARED TRANSPORTATION SYSTEM

The general composition of the discriminating function shows that 7 variables out of a possible 159 variables relating to demographic characteristics, transportation behavior and attitude are sufficient to account for approximately 92,,b of those interested in a shared transportation system from two diverse populations.

Inspection of these variables shows that demographic and attitudinal variables do not significantly discriminate people "interested" from those "not interested" in a shared transportation system. Indeed, this corresponds to previous research which shows neither type of variable to be significant influences in choice of transportation mode (e.g., Bultena, 1976).

Feasibility. The single best predictor of interest in joining a shared transportation system was perceived feasibility. The questionnaire was designed to determine how the various modifiable aspects of a shared transportation system would be designed to be feasible.

Respondents were given a series of reasons why a shared transportation system might not be feasible. The two most important reasons why a shared system might not be feasible were that respondents didn't like the idea of sharing cars (32%) and satisfaction with their current cars (27%). Acceptance of a shared transportation system is clearly based on its perception as a relative advantage over private ownership. For a large portion of the population this clearly was not the case (see Feinberg & Snuggs, 1983b for a complete discussion of this point).

The most important design attribute for individuals who believed the shared transportation to be feasible was the type of small car offered for everyday use. The larger the car the more feasible the system was perceived. And the greater the acceptability of small cars the greater the likelihood of perceiving the shared transportation as feasible. Only 16% of population believed micro-mini car to be acceptable.

The second most important attribute for individuals who found the shared transportation as feasible was the number of days the shared fleet was available for the monthly fee. Individuals who believed the shared fleet to be feasible thought they would need the shared cars significantly less often (4 days per month) than those who believed it infeasible (8 days) (p .05).

Trialability. The ability to join a shared transportation system for a trial period was the second most significant predictor of interest in such a system. Trialability affords potential participants the opportunity to see a shared transportation system as compatible with values, not complex, economically, and advantageous over private ownership (see Feinberg & Snuggs, 1983a). Indeed, survey data shows that interest in joining a shared transportation system increases dramatically when respondents are offered the option of a trial membership. In the Lafayette survey while 14% were interested in membership, 24.4/,,' were interested for a trial period (a 72- increase). Similarly, in the Columbia survey while 25 were interested in membership, 36',"' were interested in membership for a trail period (a 42,~U' increase).

Anticipated Cost for Next Vehicle and Likelihood of Purchase Within 12 Months. Although anticipated price of the shared transportation system does not significantly discriminate between those interested and not interested in membership, the likelihood of future purchase and anticipated price do. It appears that as the need and reality of purchasing a car approaches, the perceived advantage of the shared transportation system becomes crystallized. Individuals interested in joining the shared system were significantly more likely to consider buying a car within 12 months and were less willing to spend a great deal of money (see Table 2).

Compatibility with Values. Cultural values influence consumer behavior (Henry, 1976) and has shown to be significant in car choice (Rousseau, 1979) and choice of mode of transportation (e.g., mass transit, walking, car - Tan & Kundrat, 1976). Resistance to a shared transportation system obviously stems from a belief that sharing is incompatible with the widely held value of owning your own car (see Feinberg, Snuggs & Bauer, 1983). The success of shared transportation systems may depend on the ability to present such a system as being consistent with broadly held values. Unfortunate-ly, this may be a difficult task in the case of a shared transportation system.

For example, one can reasonably conceive of the car as one of the last vestiges of individual control in an increasingly uncontrollable world (Reser, 1980). The car is responsive to our needs and whims. The car intensifies the control we have and may compensate for control we do not possess. What is significant is that relative to private car ownership, transportation alternatives such as car-pooling, ride-sharing, mass transit and now a shared transportation system all lead to a sacrifice or perceived sacrifice in personal control.

Renting. Finally, individuals who have rented a vehicle in the past were significantly more likely to be interested in joining the shared transportation system than those who had not. A car rental agency, in essence, is a shared transportation system in and of itself and represents to consumers a form of the trialability discussed above.

SUMMARY

For any transportation system to function effectively, people must be willing to adopt it, use it and be satisfied by it. The direct empirical differentiation of those interested and not interested in joining a shared transportation system suggests two things. First, there appears to be a legitimate market for a shared transportation system in the U.S. Second, the success of a shared transportation system does not appear to depend solely upon objective factors which rationally could be believed to be important such as saving money and increased mobility. What appears important are subjective perceptions of feasibility and compatibility with values.

REFERENCES

Bultena, G. (1976) Public response to the energy crises: A study of citizen's attitudes and adaptive behaviors. Sociology Report #130. Department of Sociology, Iowa State University.

Feinberg, R., and Snuggs, T. (1983a) Psychological pricing and the mobility enterprise. Unpublished paper, Automotive Transportation Cente Purdue University.

Feinberg, R., and Snuggs, T. (1983b) The mobility enterprise as innovation. Automotive Transporta-_ tion Center, Purdue University.

Feinberg, R., Snuggs, T., and Bauer, S. (1983) The mobility enterprise: An innovative shared transportation system. Man-Environment Systems, 13, 87-96.

Foxx, R., and Hake, D. (1977) Gasoline conservation: A procedure for measuring and reducing the driving of college students. Journal of Behavior Analysis, 10, 61-74.

Geller, E. (1982) The energy crises and behavior science: A conceptual framework for large scale intervention. In A. Childs and G. Melton (Eds.), Rural Psychology. New York: Plenum Press.

Hake, D., and Foxx, R. (1978) Promoting gasoline conservation: The effects of reinforcement schedules, a leader and self-recording. Behavior Modification, 2, 339-369.

Henry, W. (1976) Cultural values do correlate with consumer behavior. Journal of Marketing Research, 13, 121-127.

Huberty, C. (1975) Discriminant analyses. Review of Educational Research, 45, 543-598.

Reichel, D. and Geller, E. (1981) Attempts to modify transportation behavior for energy conservation: A critical review. In A. Baum and J. Singer (Eds.) Advances in Environmental Psychology, Vol. III: Energy: Psychological Perspectives. New Jersey: Erlbaum Associates.

Reser, J. (1980) Automobile addiction: Real or imagined? Man-Environment Systems, 10, 279-287.

Rousseau, D. (1979) A cross-cultural study of automobile perception amongst three student populations. Psychologia Africana, 18, 33-46.

Tany, A., and Kundrat, D. (1976) Values and modes of travel. Perceptual and Motor Skills, 42, 214.

U.S. Department of Transportation. (1973) NationWAde Personal Transportation Study: Mode of Transportation and Personal characteristics of tripmakers. Report No. 9, Washington, D.C., U.S. Government Printing Office.