Abstract
As the consumer spatial behavior is extended, a spatial agglomeration effect of the shopping destinations is an important factor to provide for the consumer spatial behavior. When the level of aggregation of the shopping destinations influences the spatial behavior of the consumer, application of Luce model with the IIA (independence from irrelevant alternatives) property and the regularity property (for instance, Huff model and MCI model, etc.) has the possibility to become the wrong one.
As a result of the empirical analysis, it was clarified that a spatial agglomeration effect of the shopping destinations influences on the interurban consumer spatial behavior and the effect of agglomeration was different according to the characteristics of the goods. The characteristics of the goods divide into three types. One are goods where the merit of spatial agglomeration exists between the shopping destinations (“type A”). The following are goods that the demerit of agglomeration, that is, competing effect exists between the shopping destinations (“type B”). The third are goods where the effect of agglomeration does not exist (“type C”).
When the Luce model with the IIA property and the regularity property is applied to making of the consumer spatial behavior specific, the retail drawing power of individual shopping destination will be underpredicted in the goods which belong to “type A” as the level of aggregation of the shopping destinations increases. On the other hand, the retail drawing power of individual shopping destination will be overpredicted in the goods which belong to “type B” as the level of aggregation of the shopping destinations increases. The misspecification is not caused in the goods which belong to “type C” alone.
