Executive Summary
This article discusses the ramifications of the Lucas critique for marketing models. This critique says that econometric models based on historical data are often unable to predict the reaction of economic agents to policy changes because these agents will adapt their decisions strategically based on the changed policy. Although this critique may be especially relevant for marketing modelers, it has hardly diffused in the marketing literature. Limited diffusion of the Lucas critique in marketing is especially deplorable if three conditions are satisfied: (1) It is applicable to the marketing setting at hand; (2) the consequences of ignoring it are severe; and (3) ways are available that can preclude, or at least attenuate, these consequences. The article investigates to what extent these three conditions apply in marketing research.
For the Lucas critique to be applicable, the authors formulate three conditions: (1) economic agents are aware of the policy change, (2) economic agents are motivated to change behavior as a direct consequence of the policy change, and (3) economic agents are able to change their behavior. The article gives examples in which all three conditions are satisfied as well as examples in which at least one condition is not. The consequence of ignoring the Lucas critique is that marketing models yield biased predictions of the effects of marketing policy changes. However, the authors argue that the Lucas critique should be considered in light of the other model specification issues at hand. They acknowledge that the consequences of the Lucas critique are potentially severe, but they insist that this should not lead to less attention to more traditional model specification tests.
Fortunately, there are methods available to address the Lucas critique. First, the authors consider various reduced-form specifications, and several procedures (e.g., varying parameter models) that might be used to accommodate their sensitivity to the Lucas critique. Next, they consider vector autoregression (VAR) and cointegration literature, in which various tests to diagnose the Lucas critique have emerged. Third, they address structural-equations literature——most notably empirical industrial organization (IO) (new empirical industrial organization [NEIO]) models—which tries to model (and thereby eliminate) the underlying causes for the Lucas critique. The authors argue that the three modeling paradigms have different benefits and disadvantages.
The Lucas critique not only poses significant threats to the validity and practical implications of marketing models but also offers a wide range of research opportunities. Research that tries to overcome the disadvantages of the various ways to address the Lucas critique is highly desirable. In addition, a major implication that results from the Lucas critique is that the use of holdout samples and simulations in practice should be conducted with closer scrutiny and a higher degree of skepticism. There also seem to be many research opportunities into additional substantive marketing areas in which agents can be assumed to be forward looking. The authors’ hope is that marketing modelers become more aware of the Lucas critique; more motivated to test for it, as well as for other threats to validity, appropriately; and more capable of enriching their models.
Biography
Harald J. van Heerde is Professor of Marketing, Tilburg University, the Netherlands, which he joined in 2000. He holds an MSc. in econometrics and a Ph.D. (both cum laude) from the University of Groningen, the Netherlands. His work has appeared in Journal of Marketing Research, Marketing Science, Quantitative Marketing and Economics, and International Journal of Forecasting. His research deals with building econometric models in various substantive domains such as sales promotions, pricing, and new product introductions, and it is branching out into new areas such as building brand equity, loyalty programs, product crises, and price wars. In 2004, Van Heerde received the Paul Green award for best article in Journal of Marketing Research. He serves on the editorial board of the same journal.
Marnik G. Dekimpe is Professor of Marketing, Catholic University of Leuven (Belgium) and Professor of Marketing Management, Erasmus University Rotterdam (The Netherlands). He has been a visiting research scholar at UCLA and the University of South Carolina. His work has been published in Marketing Science, Management Science, Journal of Marketing Research, Journal of Marketing, International Journal of Research in Marketing, and Journal of Econometrics, among others. He has won best-paper awards at Marketing Science (1995, 2001), Journal of Marketing Research (1999), International Journal of Research in Marketing (1997, 2001, 2002), and Technological Forecasting and Social Change (2000). He serves on the editorial board of Marketing Science, Journal of Marketing Research, International Journal of Research in Marketing, Review of Marketing Science, and Journal of Interactive Marketing. He is also guest editor of a forthcoming special issue of International Journal of Research in Marketing on global marketing.
William P. Putsis Jr. is Professor of Marketing and Marketing Area Chair, Kenan-Flagler Business School, University of North Carolina at Chapel Hill, which he joined in July 2001. Prior to this time, he was Professor of Marketing and Department Chair, London Business School (1997–2003); Associate Professor, Yale School of Management, Yale University (1990–1997); and Assistant Professor, Cornell University (1988–1990). His research and academic pursuits focus on the empirical application of game theoretic models of competition, competitive strategy, the marketing of private label products, new product diffusion and product line strategy, international marketing, advertising and communications research, and sports marketing. He has numerous scholarly articles published in journals such as Journal of Marketing Research, Marketing Science, Journal of Business, Journal of Business Research, Managerial and Decision Economics, Marketing Letters, Applied Economics, Journal of Forecasting, and Review of Industrial Organization. He serves on the editorial boards of Marketing Science, Journal of Marketing, International Journal of Research in Marketing, Review of Marketing Science, and International Journal of Marketing Education.
J Marketing Research, Volume 42, Number 1, February 2005
View Table of Contents.