Assessing the Potential Effects of Differential Price Increases on Gasoline Usage

Richard C. Reizenstein, The University of Tennessee
David J. Barnaby, The University of Tennessee
ABSTRACT - This study investigates consumer reaction to various levels of gasoline price increase in terms of purchase intentions. Consumer groups formulated on the basis of propensity to consume less gasoline are comparatively analyzed. Energy policies are suggested for reducing gasoline usage and are discussed in light of varying implications.
[ to cite ]:
Richard C. Reizenstein and David J. Barnaby (1979) ,"Assessing the Potential Effects of Differential Price Increases on Gasoline Usage", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 557-563.

Advances in Consumer Research Volume 6, 1979      Pages 557-563


Richard C. Reizenstein, The University of Tennessee

David J. Barnaby, The University of Tennessee


This study investigates consumer reaction to various levels of gasoline price increase in terms of purchase intentions. Consumer groups formulated on the basis of propensity to consume less gasoline are comparatively analyzed. Energy policies are suggested for reducing gasoline usage and are discussed in light of varying implications.


For most of the twentieth century, American consumers have been accustomed to abundant supplies of relatively inexpensive energy. Products and services using these seemingly unlimited sources of power have been marketed by U. S. industry, fostering an energy affluent life style. [As an example of the relative magnitude of U. S. Energy consumption compared to other Western industrialized nations, see Helmut Becker and David J. Fritzsche, "Energy Consumption and Marketing: A Comparison of German and American Life Styles," Marketing: 1776-1976 and Beyond, Proceedings of the 1976 American Marketing Association, Kenneth L. Bernhardt, Editor, Series #39, pp. 527-532.] This almost unconscious dependency on an unending energy supply was abruptly brought into focus by the oil embargo in late 1973. Suddenly, the prospect of energy allocation, rapidly rising prices of energy intensive commodities and shortages of products and services became a reality, fostering a situation which has the potential to radically alter the American way of life. The most dramatic effects of the oil embargo were short-lived (i.e., lines in gasoline stations, curtailment of Sunday gasoline sales and quantity limitations at the pump);however, the threat of similar and even more extensive measures loom ever present on the horizon.


The importance of petroleum to the U. S. economy and its profound effects on our lifestyle make it imperative that consumer attitudes, behavior and characteristics in regard to this basic resource be studied and analyzed. It is the purpose of this research to provide insights into this important area by identifying segments with varying propensities to conserve gasoline and to describe various characteristics of these groups. This approach will hopefully enable such segments to be more easily accessed, so that conservers may be reinforced; in contrast, differential marketing tactics (pricing, promotion, etc.) may be formulated to promote increased conservation in heavy gasoline consumption groups.

Attitudinal and behavioral information related to energy has been disseminated by both public opinion organizations and academicians. Opinion Research Corporation and similar firms have been conducting periodic nationwide research since the OPEC embargo, initiated on behalf of the Federal Energy Administration and private industry (including petroleum companies) (Opinion Research Corporation, 1974).

Several studies have been published by academicians. Talarzyk and Omura primarily focused on describing consumer attitudes toward leisure-time activities in the context of the energy crisis. Findings indicated that those respondents satisfied with their leisure-time activities felt that, while their lifestyles would be affected by the energy shortage, their non-energy intensive leisure activities would continue to provide increased satisfaction. (Omura and Talarzyk, 1974; Talarzyk and Omura, 1975). Potential, rather than existing, consumer attitudes were the object of a later study by Hass, Bagley and Rogers. Results of the study indicated that increments in the perceived noxiousness or severity of an energy crisis increased intentions to reduce consumption (Hass, Bagley, and Rogers, 1975). In contrast to these cross-sectional analyses, Barnaby and Reizenstein conducted a longitudinal study of consumer attitudes toward and perceptions of the energy situation. Measures of heat, air conditioning and gasoline usage were incorporated, as was as assessment of respondent attitudes toward air pollution and pollution/energy trade-off considerations (Barnaby and Reizenstein, 1975; Reizenstein and Barnaby, 1976).

As can be seen above, some attention has been given to assessing consumer attitudes toward the energy shortage as well as gasoline usage behavior. A further key variable which has been the focus of considerable speculation is consumer reaction to potential dramatic increases in price at the pump. During the years of the Ford administration, various pricing approaches to the control of gasoline consumption were debated. These focused primarily on the deregulation of the price of gasoline but have never been implemented, given projected adverse public response and the shift in national policy brought about by the election of President Carter.

Opponents to deregulation, such as Senators Abourezk, Jackson and Metzenbaum, have hypothesized that deregulation would result in an eight cent per gallon increase in price at the pump. These individuals maintain that such an increase in a necessity like gasoline would have maximum impact on those who could least afford it: lower and middle income customers who must either use public transportation, drive long distances to work or shop, or use gasoline as an integral requirement of their occupation (i.e., traveling salesmen, independent truck operators, etc.) (Bryon, 1977).

In contrast to the above viewpoint, there is evidence to indicate that decontrol of gasoline prices would not lead to significantly higher prices at the pump in the long run; in fact, Parisi asserts that deregulation would have little long run pricing effects (Parisi, 1977). There is, further, little data to indicate that the price of gasoline is sufficiently elastic that an abrupt increase would lead to a substantial diminution in demand. In fact, there has been conjecture that gasoline may be price inelastic within certain limited ranges. Such an argument can be approached from two different perspectives: 1) Bruce-Briggs contends that even given sharp increases in gasoline price, fuel still remains relatively cheaper than in the early 1960's as a percent of U. S. family income and that 1955 price levels relative to income would be maintained for a family of median income level with gasoline selling at $.95/ gallon (Bruce-Briggs, 1974). Hence, substantial gasoline conservation would not take place without sudden large increases in price (perhaps to a dollar per gallon or more) and 2) Empirical data collected by Willenborg and Pitts indicated consumer intentions to modify gasoline consumption were high during periods of rapid gasoline price escalation. With the leveling off of the upsurge, however, the study showed that consumers appeared to rapidly adapt to higher fuel prices and revised their purchase intentions upward to former consumption levels. The authors, therefore, concluded that a relative short-term inelasticity of overall demand for gasoline existed and suggested (similar to Bruce-Briggs) that only a dramatic increase in price of perhaps 100% would be an effective conservation mechanism (Willenborg and Pitts, 1977). [It is interesting to note that while the price of gasoline continues to be driven upward by the OPEC cartel, it is in actuality rising no more rapidly than the rate of U. S. inflation. In fact, it can be seen that the recent price of gasoline at the pump as measured in 1967 constant dollars is slightly lower now than it was in 1974 during the embargo.]

In light of the uncertainties with respect to market reaction to gasoline price increases, this research attempts to identify and describe segments with varying conservation propensities at different price levels. This approach is viewed as a possible first step toward assessing the impact of incremental advances in the price of gasoline on consumer usage. It is hoped that this type of research might yield data which could provide insights into the advisability of implementing a pricing policy as a means of conserving scarce gasoline resources within unique consumer segments.


After the OPEC oil embargo had been lifted and sufficient time had elapsed to enable consumers to adapt to a less crisis-oriented view of energy consumption, a mail survey was sent to 922 residents of three medium-sized Southeastern cities, selected by means of a systematic random sample using city directories. This questionnaire had as its primary thrust the measurement of consumer attitudes and behavior with regard to energy usage, with attitudes toward environmental issues, especially air pollution, being of secondary interest. Usable responses were obtained from 390 subjects in the sample, a response rate of nearly 42%. Demographic comparisons with updated census statistics showed the sample to be generally representative of the population with the exception of a slightly upscale skew in the education and income distributions.


A variety of attitudinal and behavioral data was collected in the survey. This included energy- and pollution-related attitude and activity measures, information source variables and measures of consumer intention with regard to gasoline usage based on hypothetical per gallon increases in the price of gasoline. These variables were used to focus on identifying consumer groups exhibiting different gasoline conservation intentions with respect to the varying per gallon price increments. Once groups were formulated, multiple discriminant analyses were employed to describe segments in terms of their energy-related attitudes, use of specific information sources and selected demographic characteristics.

Prior to exploring the extent of the relationship between the set of predicators listed above and the price sensitivity groups, however, a factor analysis of the large set of attitude variables was performed to yield a reduced set of factors. These composite variables were included as independent variables in multiple discriminant analyses on groups structured according to the number of gallons per week of projected reduction in gasoline consumption.


Respondents to the questionnaire (n = 390) were asked to indicate how much they would be willing to reduce current gasoline consumption for each of nine possible five-cent increments in price per gallon at the pump. The data generated by this question are presented graphically in Figure A, with hypothesized price increases displayed along the horizontal axis and respondents grouped cumulatively along the vertical axis according to that percentage projecting conservation of a defined number of gallons of gasoline per week.

These results indicate that increases of up to $.15 per gallon at the pump resulted in 35% of the sample indicating that they would reduce gasoline consumption by at least one gallon of gasoline per week (9 percent indicated they would reduce consumption by at least 7 gallons per week). Further increases in gasoline price up to $.45 per gallon produced further decreases in consumption by up to 48% of the respondents (18 percent would reduce 1-6 gallons per week and 30 percent would reduce 7-12 gallons per week); however, 52 percent of the survey participants remained price inelastic, being unwilling to reduce gasoline usage under any of the price increases offered.

Although some caution must be exercised in accepting these data at face value given the propensity of consumers to sometimes behave in a manner at variance with their stated intentions, historical data seem to support the potential validity of the intentions measures cited in the survey. [Since the scale employed has many of the properties of a purchase probability scale, this augments the probability of predictive accuracy due to the greater sensitivity of the PPS approach in distributing responses along the intentions continuum (Juster, 1966).] Examination of such data indicates that the only real reduction in U. S. demand for petroleum occurred in early 1974, shortly after the only abrupt increase in the real dollar price of gasoline at the pump. Thus, historical backcasting seems to verify that at least a significant portion of the population is price elastic given a dramatic real increase in the cost of gasoline. (Note that in Figure A, a projected $.10 per gallon increase causes only 8% of the sample to state that they are willing to reduce gasoline consumption by 7-12 gallons per week, while a $.45 per gallon increase similarly affects 33% of the sample.)

It thus seems evident from this analysis, at least according to intentions measures, that there is a sizable group of respondents which appears price insensitive (price inelastic), while others demonstrate varying degrees of price elasticity. These results agree in part with the findings of Willenborg and Pitts showing inelasticity of demand within limited increments in price. In addition, however, the results from Figure A provide some evidence of price elasticity within smaller respondent groupings. This finding suggests that the entire gasoline market response should not be characterized as elastic or inelastic, but rather may include consumer segments possessing differing gasoline demand functions (and, conversely, different conservation functions). In order for operational promotional strategies to be derived from this conclusion, further in-depth analysis of these segments is mandatory. Group differences must be analyzed to determine the existence of segment characteristics (in this case attitudes, demographics and information sources) that are sufficiently unique to warrant the formulation of market-oriented conservation strategies and tactics for specific groups.

The objective of attempting to identify distinct gasoline conservation segments based on various consumer characteristics was met by using multiple discriminant analysis (MDA) to maximally separate groups according to a set of predictor variables. First, a factor analysis of the 45 AIO variables yielded a set of 15 factors that served as one portion of the predictors used to attempt to describe the various gasoline conservation groups. [For a more detailed exposition of the factors generated and those AIO's loading on specific factors see David J. Barnaby and Richard C. Reizenstein, "Analyzing Gasoline Consumption After the Oil Embargo: A Market Segmentation Approach," a paper presented at the ORSA/TIMS Conference, Miami, Florida, November 3-5, 1976.] In addition, both media and personal information source utilization variables and demographic measures were included as potential group descriptors.



Table 1 describes the results of SPSS multiple discriminant analysis of three consumer segments espousing different gasoline conservation intentions: reduce consumption by 0, 1-6 and 7-12 gallons per week, assuming a $.10 per gallon increase. In total, 46 different predictors, including AIO factor scores, information sources and demographics, were incorporated in the analysis. The multivariate F-ratio, significant at the .005 level, provides support for overall group discrimination; however, inspection of significant univariate F-ratios of the 46 predictors shows only 6 significant variables. [It is important to note that the greatly unequal group sizes (immediately evident in Table 1 across price increase levels) are a major consideration in this analysis, given the disproportionate group classification likelihoods associated with the unequal categories. The disproportionately small size of Group 3 (n = 25) precluded a split-half analysis to verify the reliability of the six significant predictors. Reclassification via the discriminant functions, however, always exceeded chance probability of group membership.]

Examination of the significant variable means for each of the three groups shows relatively little separation. Caution must therefore be exercised in interpreting group differences since only relative comparisons are in order (see particularly factor score mean values); however, the data do provide insights into the attitudinal and demographic composition of the three groups as well as information source importance. These data, presented in Table 1, are comparatively described for the three gasoline consumption segments in Table 2.

An immediately noticeable characteristic of Group 1 is its large size (n = 240) as compared to either of the other two groups. The significant attitudinal variables seem to indicate a lack of concern with energy- or pollution-oriented issues with respect to the other groups. In contrast, Group 3 appears to have more of an activist orientation in regard to energy/environmental issues, while Group 2 may be viewed as energy aware. This latter group has realized the potential impact of the energy shortage (although not to the same degree as Group 3); however, this segment perceives energy problems as more immediately important than pollution abatement.



Media information source comparisons indicated that Group 1, the least concerned segment, also relies least on the two significant media information sources (direct mail and radio). In addition, this group includes the greatest average number of two car families. Thus, it would appear that the least concerned segment may also be the least information-oriented segment and the least willing to give up fuel for their two cars. In comparison, although Group 3 still has a preponderance of two car families, it appears much more willing to reduce fuel consumption significantly. The fact that this group is the most oriented toward information sources, such as direct mail and radio, may be an important contributing factor toward this orientation. Given the size and orientation of Group 1 as contrasted to Group 3, it would seem as though a $.10 per gallon increase in the price of gasoline may not be an incentive to change gasoline consumption patterns in a market segment that has the least concern for the energy situation. It would appear that only those with a greater energy awareness, who are more information oriented, are sufficiently sensitive toward the problem to be willing to sacrifice gasoline consumption.

Similar analyses were performed for $.20, $.30 and $.40 per gallon price increases to determine attendant shifts in consumption and segment composition. Only the results of the $.30 price increase analysis will be shown, however, as it is at this level that the most meaningful group changes are evidenced. Beyond this point, segment composition and descriptions do not seem to change appreciably.

The data from the discriminant analysis for the $.30 per gallon fuel price information are presented in Table 3. These data are converted into verbal comparisons of the three gasoline conservation groups in Table 4. Immediately noticeable in all of these analyses is the change in size of the three groups as compared to the $.10 price increase. Groups 1 and 2 have been reduced in total by 61 respondents; Group 3 has grown commensurately larger. (Additional price increases within the limits investigated in this study did not result in noteworthy changes in group size.) These results reinforce the previous contention derived from Figure A and supported by the data in Tables 1 and 2 that a sizable number, 52% of the respondents in Group 1, continue to be price inelastic with regard to gasoline consumption. Given the above results with the attendant shift of some respondents to more conservation-oriented segments, a reexamination of the composition of the three gasoline conservation groups at the $.30 level is warranted.



A comparison of group sizes shown in Table 3 with the parallel analyses reported in Table 1 demonstrates the shift toward greater gasoline conservation with the increase in the membership of Groups i and 2. The increase in price seems to have refocused the outlook of those who do not intend to reduce their gasoline consumption (Group l) at any of the proposed levels of price increase. Whereas this segment formerly felt that the energy shortage would not appreciably hurt the economy, the potential of a $.30 gallon fuel price increase has resulted in a greater degree of pessimism and skepticism with respect to the effects of the energy situation on everyday life. For example, those in this segment now believe more than other respondents that the energy crisis will create personal hardships, while maintaining the rather skeptical view that the energy shortfall was precipitated by the oil companies. In addition, they are least confident in the fuel efficiency of the new automobiles.



It is also of interest to note that the discriminant analysis at the $.30 level yields no information source variables as significant predictors, but does provide a more distinct demographic profile of the three groups. As the hypothetical levels of gasoline price increases rise, more socioeconomic rather than information source and attitudinal variables emerge as significant discriminators. Consumer attitudes appear to have a lesser impact at the higher gasoline price levels when contrasted to demographic and situational constraints which impose boundaries on the individual's ability to react to a changing energy environment. This phenomenon can be noted by examining the Group 1 demographic descriptors in Tables 3 and 4. In contrast to the $.10 price increase level, where car ownership surfaced as the only significant demographic characteristics, the $.30 price increase profile shows the average Group 1 respondent to be older (and thus with fewer young children), less educated and less mobile than his Group 2 and Group 3 counterparts. These variables may contribute to the attitudinal rigidity described previously.

In contrast, the most conservatively oriented segment, Group 3, has enlarged considerably when compared to the $.10 per gallon price increase group (22% as compared to 6% of the total response). The attitudes differentiating this from the other groups seem to have been somewhat redirected by the additional respondents who have shifted to Group 3. This segment formerly agreed most that the energy crisis would hurt the economy; now, under the higher potential price increase, it appears to be the least apprehensive of the three groups, agreeing least that the energy crisis will cause personal hardships. Perhaps this attitude can be attributed in part to the greater degree of confidence in the fuel efficiency of the newer automobile engines. It is possible that this segment feels that the newer model autos will enable them to save a sufficient amount of fuel per mile driven, such that they can continue to drive nearly as much as they had driven prior to the oil embargo and still save an appreciable amount of gasoline. This interpretation would also partially explain the more optimistic attitude of this group in relation to the energy crisis not causing personal hardship. (This conclusion is consistent with the results of the study by Willenborg and Pitts.)

Socioeconomic variables continue to play an important role as Group 3 descriptors. In contrast to the older, less mobile Group 1, Group 3 members are, on the average, less than 40 years of age with almost half having young children. They are, in addition, somewhat more educated and much more mobile than Group 1.

Predictably, the mean descriptor values of Group 2 seem to fall between Groups 1 and 3 on all dimensions judged significant in the discriminant analysis with the exception of one variable: education. This segment is marginally the most educated of three groups, a factor which may have contributed to some conservation intentions but which by itself is not sufficient to have stimulated maximum reduction in gasoline usage.


A major use of market segmentation has traditionally been to analyze markets with respect to the consumption of goods and services. This study has demonstrated the versatility of a segmentation approach by applying it to potential conservation behavior. In addition, multivariate analyses of the different data sets have provided in- sights into the profiles of conservation group members at different gasoline price increment levels. These profiles were even further amplified by the use of several different types of descriptors: attitudinal, personal and media information source, and socioeconomic. Thus, the combination of the segmentation concept, the multivariate approach to profiling, and the incremental depth inherent in the descriptor variables, has provided insights into both segment profiles and marketing approaches designed to match the potential conservation orientations of the relevant groups.

In general, the analyses supported a division of the respondents into two broad gasoline conservation segments: 1) those who were basically price inelastic, who were non-conservation oriented and who did not intend to voluntarily change gasoline conservation practices either as a result of promotional appeals and/or pricing mechanisms; and 2) those who possessed varying degrees of price inelasticity and a conservation orientation, who might well be responsive to some combination of promotional appeals and/or pricing mechanisms. Moreover, within this latter group, a further breakdown identified subsets of consumers in regard to their conservation propensities and their gasoline consumption practices. At different price increase levels ($.10 vs. $.30), a definite shift occurred from discrimination among groups based on information source and attitudinal variables to discrimination based on attitudinal, situational and demographic variables. This seems to indicate that with greater price increases, crystallization of groups will occur and situational variables will become more critical in shaping conservation intentions. (This contention is supported by the absence of information source variables and the emergence of more situational and demographic discriminators in the $.30 analysis.) This also may imply that promotional appeals, even when targeted toward more conservation oriented segments, may be less effective in changing attitudes with greater levels of price increase, given socioeconomic constraints.



Although the results indicate that conservation intentions can be elicited from respondents similar to those in Groups 2 and 3, it is evident that the largest segment remains price inelastic regardless of the price increase within the $.05-$.45/gallon price range under investigation. Thus, promotional and pricing mechanisms dependent on voluntary consumer response appears an ineffective alternative for this group. It seems as if some form of mandatory compliance legislated at the national level may be necessary to effect substantial conservation across all segments. Over and above such uniform requirements of all citizens (such as the Carter Administration's proposed gas guzzler tax, and the required 1985 gasoline mileage standards for automobile manufacturers), market related appeals could potentially impact more conservation-oriented segments (such as Groups 2 and 3). Thus, necessary conservation could be accomplished for the most part via legislation, relying on voluntary measures for less critical contributions to gasoline conservation. Finally, legislated appeals would not only be more uniform in their overall requirements of the population, but would also undoubtedly be more successful in accomplishing necessary conservation objectives than would a strictly voluntary program to which only a portion of the population would respond.


Barnaby, David J. and Richard C. Reizenstein, "Attitudes Toward Consumption: Segmenting the Gasoline Market," Advances in Consumer Research, Volume 3, B. B. Anderson, Editor, (Cincinnati: Association for Consumer Research, 1975).

Bruce-Briggs, B., "Gasoline Prices and the Suburban Way of Life," The Public Interest, No. 37, Fall, 1974, pp. 131-136.

Byron, Christopher, "Yes, There Is An Energy Crisis," Time, Vol. 110, No. 15, October 10, 1977, pp. 61-62.

Hass, Jane W., Gerrold S. Bagley, and Ronald W. Rogers, "Coping with the Energy Crisis: Effects of Fear Appeals Upon Attitudes Toward Energy Consumption," Journal of Applied Psychology, 1975, Vol. 60, No. 6, pp. 754-756.

Juster, Thomas F., Consumer Buying Intentions and Purchasing Probability, (New York: National Bureau of Economic Research, 1966).

Omura, Glenn S. and W. Wayne Talarzyk, "Relationships Between Consumers Shopping and Leisure Time Activities and Their Attitudes Toward the Energy Crisis: A Cross Sectional Study," in Advances in Consumer Research, Vol. 2, Mary Schlinger, Editor, (Chicago: Association for Consumer Research, 1974), pp. 803-816.

Parisi, Anthony, "The Myth of Gasoline Price Controls," Business Week, No. 2466, January 17, 1977, p. 25.

Reizenstein, Richard C. and David J. Barnaby, "An Analysis of Selected Consumer Energy Environment Trade-Off Segments," in Marketing: 1776-1976 and Beyond, Kenneth L. Bernhardt, Editor, (Chicago: American Marketing Association, 1976).

Reizenstein, Richard C. and David J. Barnaby, "The Impact of the Energy Crisis on Consumer Willingness to Spend to Reduce Air Pollution," in Ecological Marketing, Karl Henion II and Thomas C. Kinnear, Editors, (Chicago: American Marketing Association, 1976).

Talarzyk, Wayne W. and Glen S. Omura, "Consumer Attitudes Toward and Perceptions of the Energy Crisis," in Marketing Contributions to the Firm and to Society, Ronald C. Curhan, Editor, (Chicago: American Marketing Association, 1975), pp. 316-322.

Willenborg, John F. and Robert E. Pitts, "Gasoline Prices: Their Effect on Consumer Behavior and Attitudes," Journal of Marketing, Vol. 41, No. 1, January, 1977, pp. 24-31.