| Multinomial logit models were used to explain consumer outlet selection when buying beef, specifically roasts, steaks, ground beef, and other types of beef. Outlets were grouped into supermarkets, butchers, warehouses, supercenters, and others, and the probability of selecting each outlet type over a range of demographic and other variables was tested. The models were estimated from household data, with 198,682 observations used in the estimation. Empirical results showed that the type of beef purchased and the size of the purchase played an importance role in the choice of outlet. Furthermore, the increase in mobility seen when consumers buy larger unit cuts could not be fully explained by price discounting. Implications for the potential growth of... |
