Benchmarking the CFS to Waybill Data

摘要

This presentation compares railroad traffic estimates from the 2002 Commodity Flow Survey (CFS) to benchmark values found in the Surface Transportation Board's (STB's) Carload Waybill Sample (Waybill). The CFS is an establishment-based survey of freight traffic conducted by the Bureau of Transportation Statistics and the Census Bureau. The 2002 CFS population estimates were developed from a 50,000 establishment stratified sample, and included all modes of freight transportation. The CFS sample frame excludes agricultural shipments from farm to first point of processing, imports from port of entry to first point of processing, crude petroleum, printed matter, commodities carried in non-revenue transportation, and shipments by government, service industries, utilities, most retailers, and households. The Waybill is an annual carrier-based survey conducted by the STB. The Waybill includes railroad-only transportation, as well as the railroad portion of rail-truck and rail-water transportation presented in the CFS. The sample frame for the Waybill includes essentially all railroad traffic. The 2002 Waybill population estimates were developed from a 600,000 record stratified sample which included 22% of all railroad traffic. The Waybill is regarded as a benchmark estimate of railroad industry traffic. Comparison of the CFS estimates of railroad traffic to the benchmark Waybill values illustrates the magnitude of in-scope issues, and identifies the differences in the railroad industry, rail-truck, rail-water, and rail-only estimates in the most recent data.