When consumers have heterogeneous perceptions about product quality,traditional parametric methods may not provide accurate marginal valuationestimates of a product's characteristics. A quantile regression frameworkcan be used to estimate valuations of product characteristics when qualityperceptions are not homogeneous. Semiparametric quantile regressions provideidentification and quantification of heterogeneous marginal valuationeffects across a conditional price distribution. Using purchase price datafrom a bull auction, we show that there are nonconstant marginal valuationsof bull carcass and growth traits. Improved understanding of productcharacteristic valuations across differentiated market segments can helpproducers develop more cost-effective management strategies.
