Explaining Variations in the Advertising & Promotional Costs/Sales Ratio: A Rejoinder

Abstract

In 1994, the authors (see Ailawadi, Farris, and Parry 1994) reported the results of a failed attempt to validate the findings of Balasubramanian and Kumar's (1990) empirical analysis of the determinants of the firm-level ratio of advertising and promotion to sales (A&P/S). Balasubramanian and Kumar (1997) have now listed several concerns with the authors’ 1994 analysis—of which two data preparation errors are valid. However, these errors have no impact on the main thesis of the authors 1994 article—market share and market growth are not good predictors of A&P/S. The authors also have analyzed currently available COMPUSTAT data, and their findings once again invalidate Balasubramanian and Kumar's (1990) results. The authors can only reiterate their earlier conclusion—neither currently available COMPUSTAT data, Profit Impact of Market Strategy data, nor brand-level data show any significant effect of market share and market growth on A&P/S. The authors conclude that Balasubramanian and Kumar's (1990) findings are simply not valid, and there is no way to identify what exactly led to their unique results because they have not made available their original data set.