Special Session Summary Does Economics Have Anything to Say About Consumer Behavior?

Russell S. Winer, University of California-Berkeley
[ to cite ]:
Russell S. Winer (1996) ,"Special Session Summary Does Economics Have Anything to Say About Consumer Behavior?", in NA - Advances in Consumer Research Volume 23, eds. Kim P. Corfman and John G. Lynch Jr., Provo, UT : Association for Consumer Research, Pages: 264.

Advances in Consumer Research Volume 23, 1996      Page 264



Russell S. Winer, University of California-Berkeley

Although the American Economics Association is one of the sponsoring organizations of the Journal of Consumer Research, very few papers using economic theory are published in the journal or presented at ACR. In addition, given most consumer behavior researchers' knowledge of economics does not go far beyond the classical utility maximization framework, there may be some question about what economics can even offer to the field of consumer behavior. The purpose of this session was to "market" economics as an alternative to psychology as a viable modeling framework for better understanding consumer behavior and to "bridge" diverse ACR constituents.



Mary Sullivan, University of Chicago

Byung-Do Kim, Carnegie-Mellon University

Brand extensions have been studied from both consumer behavior and economics perspectives. Consumer behavior researchers and economists have differed in their theoretical approaches in that the former have examined how an extension has influenced cognitive processes while the latter have studied how the extension provides information as an aid to the purchase decision. Differences between the two kinds of researchers extend to the empirical approaches: consumer behavior researchers use experimental methods while economists tend to use secondary data. Two examples of these different approaches to research in the brand extension area studies looking at the "fit" of the extension to the existing brand and the effects of familiarity and expertise on the acceptability of extensions. An example of the consumer behavior research on fit is the use of categorization theory to predict how affect associated with the product category will generalize to the extension. Alternatively, an economic approach to fit assumes that extensions are experience goods and the position of the parent brand's established product provides information about the position of the extension in the market. With respect to familiarity, consumer behavior researchers have posited that affect is more likely to be transferred to an extension by "experts" in a product category when perceived similarity to an existing brand is high. An economic or marketing science approach attempts to measure familiarity using scanner panel data by defining experience by past purchasing behavior of the parent brand. A brand choice model is then estimated linking probability of choice of the extension to this past purchasing variable. In conclusion, both consumer behavior and economics have made important contributions to understanding how extensions work.



Birger Wernerfelt, MIT

Much of the behavioral research in choice is involved with the design of experiments in which actual choice behavior violates fundamental axioms of choice. Some researchers are looking for any behavior which cannot be explained by any rational (e.g. economic) choice model. In this presentation, it is shown that some behavior which has been argued to be non-rational such as the extensive literature on context effects is, in fact, consistent with rational behavior. To do this, ways in which subjects can draw choice-relevant inferences from the set of products available in the market are described and measured. These inferences are assessed in a series of experiments, using the basic attraction and compromise designs from the literature, supplemented by direct measures of inferred product "addresses" (inferred locations of available products on a price-quality one-dimensional Hotelling line) and personal "taste addresses" (ideal points on the line). The change in choice share of a product produced by context is decomposed into three components: (1) a product address shift, (2) a taste address shift, and (3) a residual shift. What have been typically argued as violations of rational choice are instead shown to be the sum of these three effects which are based on rational, i.e. economic, assumptions.



Sridhar Moorthy, University of Rochester

In many cases, the criticisms of behavioral researchers that economic models of consumers are unrealistic are unjustified. Economic modelers have made considerable progress in reconciling seemingly irrational consumer behavior with rational models. Several examples of where economic models "do well," i.e. explain consumer behavior, are the following. One example is low involvement behavior where economists can model this as a situation where the costs of involvement (e.g. information processing) exceeds the benefits. A second example is the recent work (see presentation #2) which shows that attraction and compromise effects are actually consistent with rational economic behavior as opposed to being violations of rationality as had been previously claimed. However, there are also areas of research where economics has a problem explaining behavior. One example is preference reversals. A second example is the difficulty in explaining the frequency distribution of price endings, most of which end in either 9 or 0. A third area is advertising effects where economic explanations of advertising as having only information content. In summary, behavioral researchers need to understand what is truly psychological and what can be explained by more parsimonious economic models.



Russell S. Winer, University of California at Berkeley

The economic concepts of discounting and discount rates refer to how agents (i.e. consumers) prefer outcomes such as receipt of money sooner rather than later. Comparing discount rates between consumers allows us to infer their relative rates of "time preference" for money. For example, a consumer with a discount rate of 10% would have much greater use for the money today and value it less in the future than the consumer with the 5% rate. The concept of discounting and time preferences can also be applied to a durable goods, multiattribute context. Consumers can be conceptualized to have discount rates for attributes in that they have different rates of "impatience" for product improvements on the attributes. Thus, one can hypothesize a two-dimensional concept of attribute importance: a static concept as is usually measured by attribute importance weights, and a dynamic concept which captures consumer time preferences for the attributes. For example, in the context of laptop computers, a consumer might value the quality of the screen greater than the weight in time t (the present time), but is more impatient for improvements in weight, i.e. weight has a greater discount rate. Other relevant questions on discount rates are how they vary over product categories as well as attributes, how to model differences between consumers, and how they affect purchase incidence of different durable good categories.