Knowledge that Transforms

ance of aggregation biases, and so on. Franses (2005) provides an inventory of commonly used diagnostic tests. Franses (see Table 2 for 1998–2000 and Table 3 for 2001– 2003) also shows that there is a paucity of actual use of diagnostic tests in articles published in JMR. Such paucity may suggest that researchers have a disincentive to conduct all relevant tests. For example, it might be imagined that though researchers prefer to publish a valid model rather than one that is invalid, publishing an invalid model is still preferred to not publishing one, especially if it is difficult for readers to detect model deficiencies. Given data constraints, it is virtually impossible for researchers to accommodate all possible nuances. Thus, researchers rely on theories and experience to decide which aspects are the most critical to include in a model. With accumulating empirical evidence in the literature, the expectation is that future modeling efforts will be more informed and thus likely to provide increasingly useful (i.e., valid and reliable) results. However, I urge researchers to consult the checklist that Franses provides and to conduct all diagnostic tests when appropriate. In applied econometrics, there are three possible reasons for specific tests not to be used. First, researchers may argue convincingly that a test does not apply in the model’s context. For example, testing the null hypothesis of zero autocorrelation in the error term in a model of purely cross-sectional data is irrelevant. (Separately, I note that the use of generalized least squares rather than ordinary least squares to accommodate serial correlation in time series data is a technical correction that is not convincing unless the researcher can justify how serial correction logically arises for an otherwise correctly specified model.) Second, some tests may not yet be available for cases other than linear models and normally distributed errors. Third, researchers may argue that the violation of a particular assumption does not invalidate the substantive results. For example, consistency of ordinary least squares does not require normality of the error term. In all other cases, it is the researchers’ responsibility to conduct and show appropriate diagnostic tests. The model cannot be assumed to be valid unless proper diagnostic tests fail to reject the assumptions. Because all models are incomplete representations of reality, a central question is: How can a model that is superior to a meaningful alternative (e.g., judgment or a simpler representation than the proposed model) be obtained? If the interest of the researcher is to discover how marketing activities affect purchases or other responses (as in a “causal” model), any comparison to a model without marketing variables seems useless. Still, it might be argued that Beginning with the August 2003 issue, Journal of Marketing Research (JMR) has published one or more comments on the lead article, followed by a rejoinder. I have asked experts to provide commentary on one article in each issue that I believe has especially relevant content for researchers and managers. In the current issue, the lead article is an invited paper for which I also asked several experts to provide comments. In all cases, I provide an opportunity for the author(s) of the original article to prepare a rejoinder in the same issue. My hope is that such related reflections and commentaries on a current topic will enhance the value of JMR to readers. Although I see no reason to entice authors to express strong disagreements about specific issues, I expect that such collections of articles will enable readers to become more informed about differences in perspectives that researchers with substantial expertise and experience have on important issues.

Models of distribution channels have defined retailer and manufacturer pricing decision variables in different ways, such as absolute retail price or absolute retail margin and absolute manufacturer price or absolute manufacturer margin. This article examines whether this choice of definition affects the equilibrium outcomes from such models. It shows that the equilibrium outcomes do not change with these definitions if manufacturers are modeled as Stackelberg pricing leaders to their retailer. However, if manufacturers are modeled as Bertrand-Nash competitors to their retailer or as Stackelberg pricing followers to their retailer, the equilibrium outcomes change depending on how the retailer's pricing decision variables are defined. Moreover, if in these two cases manufacturers and retailer are allowed to define their own pricing decision variables, then (1) manufacturers are indifferent about choosing among absolute prices, absolute margins, and percentage margins, but (2) the retailer chooses percentage margins. These results have implications for both theoretical and empirical models of price competition in distribution channels.

The authors study trade promotions for durable goods, such as automobiles, for which manufacturers provide special incentives to dealers for exceeding specific sales targets. They develop a theoretical model of consumer, retailer, and manufacturer behavior based on observations about key aspects of the automobile market. The analysis provides important insights about the intertemporal effects of trade promotions and the effect of product durability on the promotion strategies of manufacturers. For example, manufacturers of more durable products benefit more from running trade promotions and give deeper discounts. The authors find empirical support when they test the theoretical results.

In electronic shopping, screening tools are used to sort through many options, assess their fit with a consumer's utility function, and recommend options in a list ordered from predicted best to worst. When the most promising options are at the beginning of the list, even seemingly advantageous factors (e.g., lower search cost, greater selection) that prompt consideration of more options degrade choice quality by (1) lowering the average quality of considered options and (2) lowering customers' selectivity in focusing attention on the more mediocre rather than the better options from the actively considered set. Study 1 shows that lowering search costs diminishes choice quality in an ordered environment. Study 2 shows that presenting consumers with the top 50 rather than the top 15 recommendations has the same effect. Study 3 shows that greater accuracy motivation in combination with lower search cost diminishes choice quality because consumers are encouraged to consider a wider range of options (lower-quality consideration sets), which ultimately leads to worse choices.

In Shiv, Carmon, and Ariely (2005) , the authors demonstrate that marketing actions such as price promotions and advertising evoke consumer expectations, which can alter the actual efficacy of the marketed product, a phenomenon they call ”placebo effects of marketing actions.” In this rejoinder, they build on the preceding commentaries and refine their framework to account more fully for factors that may influence this placebo effect, and they describe directions for further research in this new topic area.

A primary goal of research in marketing is to evaluate and recommend optimal policies for marketing actions, or “instruments” in the terminology of Franses (2005). In this respect, marketing is a very policy-oriented field, and it is ironic that so much published research skirts the issue of policy evaluation. Franses’s article draws much needed attention to the question of what sort of model is usable for policy simulation and evaluation. Our perspective on what constitutes a valid model for policy evaluation differs from Franses’s view, but we believe our view complements his in many important respects. We also strongly believe that marketing has much to contribute to the literature on structural modeling. We outline some of what we believe are the advantages for marketing scholars of using structural modeling for policy evaluations and some of the challenges presented by marketing problems. Franses focuses on a reduced-form sales response model in which the outcome variable (yt) is modeled conditional on marketing variables (xt). If customers anticipate future marketing actions and take these into account in responding to the environment at time t, an additional equation is appended to the system to describe the evolution of the xt variables. In Franses’s view, this system can be used for policy simulation if both the y and x equations have timeinvariant parameters. That is, the Lucas critique, which implies that parameters of reduced-form models change if the policy regime changes, does not apply. According to Franses, a model must pass standard diagnostics, possess good predictive properties, and exhibit parameter stability to be useful for policy simulation. We applaud the attention Franses is bringing to model diagnostics. We believe that structural work in both marketing and economics should pay close attention to the central features of the data. Increased use of model diagnostics will help ensure that structural models are capable of capturing these features. However, we do not believe that all the criteria proposed by Franses, such as out-of-sample validity and parameter stability, are either necessary or sufficient to render a model useful for policy simulation. Reduced-form models can pass all diagnostics, including out-of-sample validation, and still provide misleading predictions about the effects of policy changes. Reduced-form

The marketing literature has addressed the issues of heterogeneity and endogeneity when estimating a choice model with household-level panel data. When using these data at the stockkeeping unit or the Universal Product Code level, choices for each item in each of the time periods under consideration cannot be observed. Without such information, it is difficult to control for item- and time period–specific unmeasured characteristics because there is no information on alternatives during those periods in which they are not purchased by any of the panelists. In general, when a product category has many alternatives, each with fairly small shares, the household sample may not contain sufficient choices for each alternative, thus negatively affecting the ability to control for endogeneity with household data. In contrast, because aggregate store-level data (for those stores in which the panel makes purchases) are the aggregation of purchases by all households visiting the stores, the data contain the time period–specific item-level information required to account for endogeneity as long as each item has some sales in each time period. Given the relative merits of household data to estimate the distribution of heterogeneity and store-level data to address the endogeneity problem, the authors propose an integrated estimation procedure that uses the information in both sources. They provide empirical results from their model using data on the fabric softener market. They extend their approach to situations in which there is variation in purchase quantities that households choose.

The Lucas critique has been largely ignored in the marketing literature. We present a number of conditions under which the critique is most likely to (also) apply in marketing settings. Next, we provide some perspectives on how to diagnose and accommodate the Lucas critique, and identify various avenues for future research.

In this article, the author investigates the impact of a brand extension's success versus failure on customer evaluation of brand counter-extensions. A counterextension is a brand extension that is launched into Category A by Brand 2 that belongs to Category B in a reciprocal direction to a launch of a previous extension into Category B by Brand 1 that belongs to Category A. The results from five studies show that customers evaluate a counterextension more favorably when the preceding extension is a success rather than a failure. Furthermore, the evaluation of the counterextension is superior if it is launched by a major brand, especially if the previous successful extension was also launched by a major brand. Finally, a successful extension indirectly dilutes a brand and results in a greater loss in choice share to a counterextension than does a failed extension. The key findings generalize to a sequence of extensions across more than one intercategory boundary.

Retailers typically engage in some form of price discrimination to increase profitability. In this article, the authors compare the impact on retailer profitability of two price discrimination mechanisms: quantity discounts based on package size (second-degree price discrimination) and store-level pricing or micromarketing (third-degree price discrimination). Whereas the latter has been well addressed in the marketing literature, there is limited empirical research on the use of quantity discounts for price discrimination. Using store-level sales data, the authors estimate a structural demand model, accounting for parameter heterogeneity and price endogeneity. They combine the parameter estimates with a model of retailer pricing to conduct optimal pricing and profitability simulations under several scenarios, ranging from constraining the retailer not to engage in any form of price discrimination to the least restrictive scenario of setting nonlinear price schedules specific to each store. The pricing simulations enable the decomposition of profitability as a result of the different forms of price discrimination. Profits are greatest when retailers combine second- and third-degree price discrimination. The authors find that the ability to engage in second-degree price discrimination contributes more to retailer profitability than does third-degree price discrimination.