Author: Glyn Wittwer
The objective of this study is to assess the suitability of Commodity Flow Survey (CFS) data released by the US Census Bureau as a check on the estimates of inter-regional trade generated in creating the USAGE-TERM master database. A close inspection of the Census Bureau's CFS data indicate that they record movements to and from transport nodes. In some cases, transport nodes may align with production origins and final use destinations. In other cases, nodes appear to be intermediate points rather than origins or final destinations. This implies that CFS data are often incompatible with the trade flows in a CGE database. Merchandise, that is primary and manufacturing commodities, account for no more than 15% of GDP in the U.S. economy. Therefore, even comprehensive merchandise trade flow data would have limited use in a CGE database. The usefulness of the CFS data is diminished further by its concentration on bulky goods, which account for a small fraction of total trade flows. Bulky trade flows may account for a substantial proportion of the volume of trade but make a small contribution to total economic activity. Mining products excluding oil and gas account for 50.9% of the recorded weight in the survey, but just 3.9% of the value of trades – and only 0.3% of GDP. The CFS data might be useful for examining transport logistics but are of little use in CGE database preparation. There is no evidence that the CFS data supersedes the Horridge gravity method of allocating inter-regional trades. However, CFS data point to the desirability of noting the difference between transport in the Mississippi basin and elsewhere. The basin relies heavily on water transport for moving agricultural, mining and fuel products.
JEL classification: R11, R13, R15, C68.
Keywords: Regional CGE modelling, inter-regional trades.
Working Paper Number G-275 can be downloaded in PDF format. To print this you will need the Adobe Acrobat Reader.
Go to working papers page