A methodology to estimate current commodity inflows to a substate region using a variant of the standard regional input-output model and commodity flow data from the 1993 Commodity Flow Survey is pre- sented. Existing data in the Commodity Flow Survey do not go below the state level, making the estimation of commodity flows to a particu- lar substate region difficult. By combining state-level commodity flow data with a “supply-side,” commodity-by-industry input-output model, an estimate of commodity flows to a region can be carried out based on the region’s existing industrial structure. The information on current commodity inflows is particularly useful for analyzing the total poten- tial demand that might be captured by a new mode or service. For exam- ple, one application could be estimating the potential for air cargo shipments to an area that is not currently served by local air cargo. A typical problem faced when assessing the feasibility of new air cargo service to a region is estimating the potential demand for the service. This potential demand can be placed into two categories. The first, air cargo outflows volume, is relatively straightforward to estimate, given adequate data on the region’s industrial structure— one can infer the level of regional output and, based on simple esti- mates of export-orientation to outside the region, gauge what might be the potential demand for regional air cargo service by these exporting industries. The estimation of the second category, air cargo inflows, is considerably more complicated. The main reason is that direct data on commodities being imported to a substate region are not available. While the estimation of some commodity inflows can be tied to households and their consumption patterns, the estimation of com- modities for industrial and commercial use is not estimated so eas- ily. Existing data on commodity flows, in particular those contained in the 1993 Commodity Flow Survey (1), are rich in commodity- level detail but are not disaggregated below the state level. Estimat- ing commodity inflows to a substate level, particularly if the region accounts for a relatively small proportion of a state’s economy, requires indirect estimates. Without information on the actual lev- els of commodity inflows to a region, it is difficult to predict what might be the market share of these inflows for a new regional cargo service. A possible solution to the problem of estimating actual commod- ity inflows to a region could be through the estimation of a gravity- type model. Gravity models have been applied extensively to the analysis of passenger trip generation, and examples exist of their application to freight demand modeling. A gravity model for freight would take the following general form: T fS D C ij k i k j k ij k = ( 29 , , TRANSPORTATION RESEARCH RECORD 1653 Paper No. 99-1093 17 where T k ij = shipments of commodity k between regions i and j, S k i = supply of commodity k in region i, D k j = demand for commodity k in region j, and C k ij = generalized cost of shipping commodity k between i and j (2). The gravity model is essentially estimating the importance of fac- tors that “attract” freight inflows to a sample of areas. In terms of predicting commodity inflows to a substate area, this model could be estimated at the state level, and the estimated parameters could be used to predict inflows to a substate region. However, the data requirements to calibrate this type of model are significant, particu- larly with respect to the shipping costs between states. This same conclusion applies to other, closely related models based on discrete choice analysis, including disaggregated freight generation models estimated using the common logit framework. An alternative approach is proposed, one that bases estimates of actual commodity inflows to a substate region entirely on the region’s industrial structure. The approach relies on input-output data, using information on the input requirements of industries to estimate regional commodity inflows. In a form that is less famil- iar to users, namely in the commodity-by-industry format, input- output data are used to share state-level commodity inflows to substate regions by directing these inflows to the industries that use them as inputs. Further, the procedure can be carried out fairly easily, relying entirely on published national input-output data, existing state-level commodity flow data from the 1993 Commod- ity Flow Survey, and regional data on employment or earnings by industry. The procedure described below involves two steps. First, using a modified regional input-output model, one defines the proportion of various commodities that are used by various industries in a region of interest. Then, one applies these proportions to existing state-level commodity inflow data from the 1993 Commodity Flow Survey to “share down” the state-level flows to the region. As will be shown, the methodology takes into account the possi- bility that the input needs of a regional industry are met, in whole or in part, by regional suppliers. By accounting for existing patterns of regional interindustry freight flows, the accuracy of estimated regional freight inflows is greatly increased. This aspect of the methodology contrasts with the approach suggested in a 1983 report by Memmott (3). While also based on input-output models, his sug- gested procedure for estimating regional freight flows does not account for the possibility of freight inflows being supplied region- ally. As a result, the applicability of the approach for accurately estimating inflows from outside the region is limited. Estimation of Commodity Inflows to a Substate Region An Input-Output Based Approach PIERRE VILAIN, LOUIE NAN LIU, AND DAVID AIMEN Louis Berger & Associates, Inc., 100 Halsted St., East Orange, NJ 07019.