Endogenous Modeling of Late Entry Penalties for Packaged Goods By Gurumurthy Kalyanaram

This editorial is abstracted verbatim from the following research manuscript: “ENDOGENOUS MODELING OF LATE ENTRY PENALTIES FOR PACKAGED GOODS by Gurumurthy Kalyanaram and Glen L. Urban, MIT Working Paper (WP #3412-92MSA).” The working paper can be obtained from Dr. Gurumurthy Kalyanaram.

Introduction

Numerous previous papers have documented the empirical relationship of early entry with market share advantages generally using a single equation model. However, there are two major threats to the estimates from the single-equation models arising from possible endogenous entry phenomena. First, the marketing variables may themselves be a function of order of entry. For example, the level of advertising may be systematically lower for later entrants. If this is not modeled, “innate” order effect may in fact be the structure of spending of later entrants and not a true market penalty. Similar arguments apply to price, promotion and distribution.

The second threat is that entry itself may be endogenous. Vanhonacker and Day (1987) discussed this as an example of an area where reserve regression is relevant. Moore, Boulding, and Goodstein (1991) explicitly suggest that entry may be a function of the skills and resources of the firm. When this is true, they show that the underlying order effect will be biased and may even be of opposite sign. They do not build an endogenous model for order of entry effects, but they show the dangers of ignoring the effect in estimation. Similarly, while Rodrigueze-Pinto, Gutierrez-Cillan and Rodriguez-Escudero (2007) did not study the endogeneity issue, they found that the entry effect is moderated by marketing and R&D resources. These findings make a compelling case for this paper.

The issue of endogeneity, while it has not been addressed in substantial measure in the context of order of entry effect, endogeneity has been studied well in the context of advertising and it has been found that advertising should be modeled as an endogenous decision (Dave and Saffer 2009, Kalyanaram 2009, Chintagunta, Kadiyali and Vilcassim 2006, Cotterill, Putsis and Dhar 2000, Villas and Winer 1999, Erickson 1992, Pindyck and Rubinfeld 1991).

Accordingly, we develop in this paper an explicit simultaneous equation model that includes the endogenous effects of order of entry on the levels of marketing variables and the entry order itself. The purpose of this paper is to examine the endogenous effects that could change the estimated magnitude and significance of the order of entry penalties. We also incorporate the dynamics of growth which is relevant and important in the context of consumer package goods.

The structure of this paper is as follows. We briefly review the literature, and then propose a structural equation model that examines the endogenous effects of order of entry on marketing variables and the possibility that order itself is endogenous and a function of the skill and power of the firm. In the following sections, we present the empirical results of applying the new model to the original data, examine this existence and magnitude of the innate order of entry effects, analyze the sensitivity of the structure, and close with an identification of directions for future research.

Read Full Article