Advances in Consumer Research Volume 11, 1984 Pages 633-635
PRICE EFFECTS ON CONSUMER BEHAVIOR: A STATUS REPORT
Jerry F. Conover, University of Arizona
The role of price in consumer behavior is certainly both influential and complex. As the four papers in this session demonstrate, this role can be explored in many ways, each leading to different insights about how price influences consumer decisions. The review that follows highlights some of the key issues and problems suggested by these papers, offering ideas for the further development of research and theory on consumer response to price.
"ENTRY/EXIT DEMAND ANALYSIS"
Dickson and Sawyer's (1984) exposition of "Entry/Exit Demand Analysis" presents a well documented review of previous attempts to assess a managerially important relationship: the response of consumer demand to various price levels for a product. As the authors point out, there are more troublesome limitations to those earlier approaches to estimating demand curves, and entry/exit demand analysis represents a technique for overcoming those difficulties.
Among its advantages, Dickson and Sawyer note, is the technique's versatility of administration. It can be administered in person, in group settings, or by mail, which makes the approach valuable to pricing decision makers in a variety of situations. Moreover, the technique elicits information about all the competing brands, and thus gives managers more useful data on which to base price-setting strategies than do demand-estivation techniques that refer to generic product categories.
On the other hand, entry/exit demand analysis does have some inherent drawbacks that limit its utility. Perhaps most significant is the requirement that subjects be familiar with the brands being considered. In order to give meaningful answers to such questions as, "At what lower price would you buy Brand A?" the consumer subject must have some knowledge of Brand A's characteristics. Thus, the technique would be of limited value for estimating demand for new or otherwise unfamiliar products. A related problem is the laboriousness of collecting adequate data when the choice set is large (as is common with many consumer packaged goods). This problem might be overcome, however, with appropriate sampling of brand subsets to which each subject could respond, pooling results from different subjects together. Even so, the problem of unfamiliarity with some brands could confound efforts to obtain a balanced set of responses for all the brands.
In addition to these concerns with administering the measure, certain demand artifacts might complicate interpretation of its results. The whole thrust of the questioning focuses the subject's attention on price, and thus may lead to small price differences causing larger responses than they would in an actual marketplace situation. This, however, is an empirical question that could be addressed by comparing demand curves obtained through entry/exit analysis with corresponding curves from field experiments manipulating prices. (Such field experiments would also serve the valuable purpose of validating the general usefulness of this technique.) One other possible bias in subjects' responses might result from the use of odd-prices as the starting points for the various brands' prices (as was the case in the example provided in this session). Although those prices were selected to represent actual market conditions. they may lead subjects to specify entry or exit prices in terms of odd-prices, whereas in a real choice situation, they might enter or exit the brand's market before an odd-price level is reached. Thus, the precision with which entry/exit analysis identifies break points may be compromised. Despite these potential disadvantages, entry/exit analysis represents an important step forward in the development of a practical and meaningful basis for estimating price-response functions.
In his analysis of market response in situations where price is used as a cue to quality, Shugan (1984) presents a theoretically sound logical development of factors influencing price-quality relationships. Moreover, this conceptualization of market response as influenced by different segments of consumers, each with different perceptions of and preferences for product characteristics, leads to a managerially useful means for determining the optimum levels of price and quality to set for a product.
In order to apply Shugan's model to a real market situation, one would need data about the market's response coefficients for both price (i.e., "b" in Shugan's equations) and quality ("c"). The coefficient of sensitivity to changes in price could be determined from data collected in actual field manipulations of prices, or from experimental simulations, such as Dickson and Sawyer's (1984) approach. The quality-response coefficient, however, would be somewhat more difficult to determine empirically. Manipulations of product quality would, in many cases, be rather impractical, requiring changes in production processes for small-scale testing. levels of quality might be described verbally for some product attributes, but for others that are difficult to describe explicitly (eg., such subjective characteristics as "styling" or "luxuriousness"), it would be hard to experimentally test quality-response via simulated choice conditions. Thus, the practical applicability of this model may be limited to products characterized by relatively simple, easy-to-describe features.
Such questions of application notwithstanding, Shugan's formulation does have considerable theoretical appeal. It represents a straightforward. yet elegant account of the determinants of price-quality relationship in consumer purchases. From this initial formulation, further refinements of the model might be in order to adapt it to conditions typical of the real world. For instance, in its present form, the model assumes that consumers may be sensitive to either price or to quality, but not to both. Realistically, however, many consumers may value both low price and high quality, and thus may have to make tradeoffs in their choices. How might the model be extended to account for such consumers?
Another type of consumer who is currently unaccounted for in the model is the person who simultaneously seeks more than one quality from a product. Such a consumer is likely typical for a large number of products, and the model could probably be adapted to handle such consumers' existence.
Given adaptations such as these, Shugan's model of market response to price and quality may well provide managerially useful insight into product demand, as well as offering a conceptually appealing account of why price-quality relationships occur.
"PRICE AS A 'GIVE' COMPONENT"
Ahtola-(1984) presents a view of price that is rather different from prior conceptualizations. In this "exchange-theoretic" model, price is seen as a common (and perhaps the major) element on the "give" side of a give/ get equation that a buyer faces. That is, one gives up a certain price in return for a bundle of attributes that constitute the "get" package of the transaction. If one adopts this point of view, then Ahtola is correct in stating that the overall value of the "get" package should be measured without regard to a specific price, but rather with respect to the value of "not getting" it.
The propositions of Ahtola's model present some interesting research questions concerning situations in which the predicted outcome is not clear. For example, two assumptions are stated in Ahtola's paper:
1. If two attributes at a low level of the attribute hierarchy are seen as interacting, then they are not evaluated separately by the consumer; instead, assessment of their value occurs at the superordinate level.
2. On the other hand, if separate prices are known for each attribute (as might be the case for automobile options or pizza toppings), the consumer will evaluate each attribute separately.
It is easy to imagine a situation where these two propositions conflict, when prices are known for interactive lower-level attributes (for example, when the value of getting anchovies on one's pizza depends on whether one also orders jalapeno peppers on it, and the prices for both toppings are known). What, one wonders, would happen in the evaluation process then? Would the value of each attribute be determined separately, or would a summary evaluation of the overall attribute-six occur?
In addition to suggesting research questions based on the present model, this work suggests extensions of the model. As it now stands, Ahtola's model describes a consumer evaluating a given choice option; that is, whether to buy d given brand or not. Further development of the model might profitably address the question of choice among a set of brands. That is, if a consumer declines one alternative to "get" another, how much will he be willing to "give" up? Does the evaluation of one brand depend on what must be "given up" by not selecting another? And in what way? Such elaborations of the model are encouraged.
"CONSUMER RESPONSE TO PRICE"
Zeithaml's (1984) review of the literature on psychological response to price information has long been needed. is research area is characterized by a hodge-podge of concepts and measures, many of which overlap each other, and the boundaries of which are often indistinct. Zeithaml's call for refinement of our concepts and measures is very well taken. Without the sort of work she encourages, our understanding of consumer response to price is bound to progress very slowly.
This is the first published work to systematically explore the past constructs and measures of price response, and to tie them together through the use of a common conceptual schema. Although the review of this literature is more descriptive than critical, it does help organize this body of work so that future research can better select appropriate measures with an eye toward establishing some continuity of usage of terms and operationalizations. If future work in this area does achieve some consistency in its measures, then the literature will become more amenable to techniques for assessing the general conclusions to be drawn from the entire body of research results (eg., meta-analytic techniques; see Class, McGaw & Smith 1981, and Reilly & Conover 1983 for examples).
The Jacoby and Olson (1977) framework for studying psychological response to price is an appropriate basis for organizing this literature. The task of classifying a given concept or measure in.to the categories of this schema, however, is not always straightforward. In particular, the stages of forming perceptions of prices, and forming attitudes toward prices are often confusingly similar. For example, when such concepts as "price acceptability" or "perceived value for the money" are classified as attitudinal responses, they seem to share much in common with concepts such as "perceived savings," which is described by Zeithaml as perceptual response. In both cases, the consumer is making judgments about a given price relative to other prices (i.e., savings relative to prior or usual prices; value-for-money implies more standard of comparison).
Another classification of variables as "attitude toward P-price" also seems somewhat dubious. The consumer's "difficulty in. evaluating prices" (Nystrom, Tamsons & Thams 1975) and "confidence in price knowledge" (e.g., Zeithaml 198.2) seem to be not so much measures of attitude toward price as they are self-reports of the consumer's processing abilities. That is, they are not really responses to prices, per se, but rather to the task of processing price information. Perhaps these concepts should not be classified in. the Jacoby and Olson (1977) schema along with all the others reviewed by Zeithaml.
The concept of "price image," referring to the perceived level of prices in ore store relative to other stores, is perhaps another concept that doesn't fit neatly into this overall schema. The majority of concepts reviewed by Zeithaml reflect responses to prices of individual products, whereas "price image" refers to a global assessment based or perceptions of numerous products' prices (or possibly on other sources of information such as store advertisements). In the former case, consumer reactions to prices are caused by discrete items of O-price, whereas "price image" is a response to less distinct stimuli. More work is needed. to establish the parameters influencing consumer assessment of store price images and how those images affect processing of individual product prices within the store.
The "Integration of P-Price with Other Information" classification is one which might be extended to include additional concepts besides the relative salience of price vis-a-vis other attributes. For example, in evaluating a given price at the time of a shopping decision, the consumer must integrate that price information with price knowledge accumulated from prior experience and from other information sources, in order to interpret it effectively. And in making "value for money" judgments, the price must be assessed relative to other product information (e.g., the "get" package, Ahtola, 1984).
The difficulty in placing different price-response concepts neatly into a single schema stems from the lack of such a schema in the original development of those concepts. In. future research, however, a coherent framework for conceptualizing consumer response to price will greatly facilitate progress in this field. One particularly interesting area for such investigation is the topic of how P-price is encoded and stored. For example, it would he useful to know what determines the accuracy of price knowledge, and whether that accuracy is influenced by market conditions, situational factors, or individual differences among consumers. Choosing the appropriate measure for studying price knowledge accuracy depends on just which aspect of it one wishes to explore. The "exact price recall error" measure discussed by Zeithaml (1984) appears to offer a conceptually appealing formula for assessing the error in one's estimate of a product's price-although there may be occasions in which it would he equally useful to study the direction of that error as well as its magnitude. On the other hand, if one is concerned with knowledge of relative prices, the best measure to use is less apparent. Is Zeithaml's earlier work (1982), "price comparison error" was operationalized by counting the differences in rank positions between the correct rankings of brands within a category and the subject's estimated rankings. Such a measure poses difficulties in comparing categories with different numbers of brands, or even comparing subjects within a category if they do not know the relative prices of more brands (and, thus cannot rank some brands without sheer guessing). A measure more amenable to such comparisons, but based on the same sort of task, would involve calculation of the correlation between estimated brand ranks and actual ranks. Or one right consider using a different sort of task, especially if the focus is on knowledge of the relative price of a single brand within its category, not on the relative price of every brand in the category. For such a question, the researcher right ask the subject to estimate on a scale (ranging from "least expensive brand" to "most expensive brand") where a given brand fits in price. As can be seen, even the issue of investigating the accuracy of a consumer's price knowledge is not without its problems when it comes to selecting appropriate measures.
In summary, it is readily apparent that there is much to be learned about the influence of price on consumer behavior. The papers presented in this special topic session represent rather different approaches to studying this important area. As Zeithaml (1484) points out, the benefits to be gained by pooling knowledge from a wide variety of independent studies will be greatly expedited to the extent that we who carry out the research adopt consistent concepts and measures for our work. Through the development of standard price-response concepts and techniques for studying them, and with extensions of prior conceptualizing such as those represented by the papers of this session, advancement of knowledge of consumer price response will surely follow.
Ahtola, Olli T. (1984), "Price as a Give' Component in an Exchange Theoretic Multicomponent Model," Advances in Consumer Research, 11, forthcoming.
Dickson, Peter R. and Alan C. Sawyer (1484), "Entry/Exit Demand Analysis," Advances in Consumer Research, 11, forthcoming.
Class, Gene V., Barry McGaw and Mary Lee Smith (1981), Meta-Analysis in Social Research, Beverly Rills, CA: Sa&e Publications.
Jacoby, Jacob and Jerry C. Olson (1977), "Consumer Response to Price: An Attitudinal, Information-Processing Perspective," in Y. Wind and M. Greenberg, eds., Moving Ahead with Attitude Research, Chicago: American Marketing Association, 73-86.
Nystrom, Harry, Hans Tamsons and Robert Thams (1975), "An Experiment in Price Generalization and Discrimination," Journal of Marketing Research. 7 (May), 177-181.
Reilly, Michael D. and Jerry H. Conover (1984), "Mete-Analysis: Integrating Results from Consumer Research Studies " Advances in Consumer Research, 11, forthcoming.
Shugan, Steven M. (1984), "Price-Quality Relationships, Advances in Consumer Research, 11, forthcoming.
Zeithaml, Valarie A. (1982), "Consumer Response to In-Store Price Information Environments," Journal of Consumer Research, 8 (March), 357-369.
Zeithaml, Valarie A. (1984), "Issues in Conceptualizing, and Measuring Consumer Response to Price," Advances in Consumer Research, 11, forthcoming.