【摘要】This paper presents a new method for estimating regional trade flows using transportation survey data describing commodity origin-destination flows. Explicit attention is paid to the difference between commodity flows and trade flows that arises from the presence of transhipment points. Observed commodity flows in the transportation survey data are converted to production-consumption trade flows that are consistent with the multi-regional input-output framework. Regional trade flow estimates are then reconciled with regional production and consumption estimates using a mathematical program that aims to make minimal adjustments while imposing known multi-regional input-output accounting identities. It is shown that commodities originating or terminating at a transhipment point should be reassigned to their probable production origins or consumption destinations as long as an unbiased sample of previous observations is available. As the number of observations increases; the prediction error of the production origin or consumption destination decreases exponentially. A real-world case study in the Province of Ontario in Canada demonstrates the feasibility of estimating interregional trade flows from commercial vehicle survey data and shows that the estimated pattern of trade flows is maintained after adjustments are made to satisfy accounting constraints. Therefore; it is possible to create a balanced multi-regional input-output model based on a commercial vehicle survey and the limited supplementary data available at the regional level.

【文献来源】Bachmann C;Kennedy C;Roorda M J. Annals of regional Science.2015（3）