Due to its internal consistency and ability to account for spatial variables like the proximity and size of alternative markets, the Trade Flow method is presumed to be superior to the Econometric method for estimating Regional Purchasing Coefficients (RPCs). For this reason, it is the default and recommended method of trade flow estimation for all state- and county-level models. For the US total (where domestic interregional trade estimation is not required), RPCs are derived using Supply/Demand pooling. Zip Code level RPCs are derived using the Econometric RPC method as trade flow data is not currently available at the ZIP Code level.
US Total File RPCs are derived using Supply/Demand. A commodity's RPC is calculated as local supply of that commodity divided by local demand for that commodity, capped at 1.00. This assumes that local demand is completely satisfied by local production to the extent possible. As a consequence, there are only exports of a commodity if local supply of that commodity exceeds local demand for that commodity. This implies that there is no "cross-hauling". Cross-hauling occurs when a region both exports and imports a particular good or service. Cross-hauling can arise from the product-mix issue, whereby a sector produces a variety of products. For instance, a region may export apples and import mangoes. Most I-O models will consider apples and mangoes to be the same commodity, namely "fruit", and thus will both export and import that commodity. Cross-hauling of a commodity can also arise from brand loyalty, long-term contracts, etc.
When cross-hauling is ignored, interregional trade is underestimated while RPCs and regional Multipliers are overestimated. Thus, the Supply/Demand Pooling method is not ideal for estimating commodity RPCs at the sub-national level. However, it is ideal for estimating RPCs at the national level. At the U.S. level, there is no domestic trade and no need to estimate the regional trade flows. We only need to know the proportion of U.S. gross demand for each commodity that is met by U.S. suppliers. We have data on U.S. foreign exports of each commodity; the remaining supply must therefore go to domestic consumption. The Supply/Demand Ratio is thus domestic supply divided by gross demand.
Econometric RPC (E-RPC)
Like the Supply/Demand Pooling method, the E- RPC methodology also only estimates RPCs; it does not estimate trade flows between regions, only the proportion of local demand that is met by local producers. However, in contrast to the Supply/Demand Pooling method, E-RPC allows for the possibility of cross-hauling. IMPLAN's Zip Codes and Congressional District data uses the E-RPC trade data method.
With this method, RPCs are derived by econometric equations that are estimated using the characteristics of the region. There is a different equation for each commodity with variables filled by study area data. The RPCs are limited by the Supply/Demand pooling ratio; local use of local supply cannot exceed local supply.