2. Methodological Issues
The theories on the effect of market concentration on product variety, discussed
above, suggest that the new product introductions can either promote competition or deter
spatial entry. Hence for the purpose of empirical estimation of the effect of market
concentration on product variety, it implies that market concentration is endogenous as
market concentration gets affected by product introductions. Empirical evidence also
suggests endogeneity of market concentration as an explanatory variable for product
introduction. Zellner’s (1989) study shows that new products introductions and market
concentration have significant positive effects on each other in a simultaneous equations
system6. Geroski and Pomroy (1990) show that product variety affect the evolution of
market concentration using data from UK manufacturing industries. Hence the empirical
studies should test for potential endogeneity of market concentration while explaining
product variety. Failure to account for the endogeneity of market concentration can
produce biased results. With the exception of Zellner (1989), all the other studies on food
industry treat market concentration as exogenous.
The theories and the empirical evidence from Geroski and Pomroy’s (1990) study
suggest a dynamic relationship between industry concentration and product variety.
Hence a panel data set is more appropriate to analyze the interactive relation between
these two variables of interest. Yet, all the empirical studies on the food industry, with the
6 This study on US food industry belongs to the structure-conduct-performance (SCP) strand of literature
and focuses on the use of advertising as a substitute strategic tool for new product introductions. The cross
sectional analysis based on new products introductions data for 1977-78, used a simultaneous equations
system with advertising intensity, price-cost margin, concentration, and new product introductions as
jointly endogenous.