ACCOUNTING FOR HETEROGENEITY AND NONSTATIONARITY IN A CROSS-SECTIONALMODEL OF CONSUMER PURCHASE BEHAVIOR

When calibrating a brand choice model cross-sectionally, a measure of brand loyalty is often introduced into the utility function to account for differences in utility across households and over time. One of th e most widely used measures of brand loyalty, proposed by Guadagni and Little (1983), is an exponential smoothing model of past choice behav ior by the household. In this study, we argue that the exponential smo othing model of brand loyalty cannot properly distinguish between sour ces of variation in utility due to heterogeneity (across households) a nd sources of variation due to nonstationarity (within household over time). We introduce a new measure of brand loyalty, derived from a non stationary Dirichlet-multinomial choice model, in which heterogeneity and nonstationarity are handled distinctly.