Measuring Technology and Pricing Differences in the Digital Still Camera Industry Using Improved Hedonic Price Estimation

Abstract

In this paper we present an improved hedonic price regression for cases in which many implemented technological characteristics are to be included into a set of explanatory variables. This approach modifies Massy’s (1965) principal components regression in which original dummy explanatory variables corresponding to the characteristics are transformed into principal components. The selected subset which we obtained from the principal components regression was further orthogonally rotated by Kaiser’s (1957) varirnax criterion in order to reconstruct explanatory variables with a simple structure for the final hedonic regression, resulting in a more straightforward understanding of hedonic regression in the case of many explanatory variables. This approach is applied in order to compare different pricing strategies in the digital still camera industry, where the products contain 35 different technical attributes. The pricing strategies were measured using relative errors: differences between log actual prices and log quality-adjusted prices were identified using hedonic regression. The result shows that those companies, which have priced their products cheaper than the value of contained technologies, or than the prices of competitors, seem to have gained a larger market share.