Abstract
Questions relating to the optimal structure of state-substate multiregional models are addressed by comparing four multiarea models estimated for Arizona and its subcomponent areas. The study arrives at three major findings on structural methodology. First, greater accuracy is achieved at both the state and substate levels when the areas are estimated by a separate single-area free-standing model. Second, the multiarea model structures which maximize accuracy at the state level also
Get full access to this article
View all access options for this article.
