The materials cost deflators are calculated using data from the use-make (input-output) tables and the GDP-by-Industry data of the Bureau of Economic Analysis (BEA). The raw data are available at the BEA website ( http://www.bea.gov) under the Industry tab. The use-make tables disaggregate each industry's total materials cost into the amount spent on each specific material. The share of each material in the industry's total materials cost is then used to weight each material's price index, resulting in a weighted-average price index for the industry's total materials cost. The benchmark versions of the use-make tables, which are generated every 5 years, are used. A moving average of the weights from the two surrounding benchmark years is used to interpolate the weights for the use-make tables in the intervening years. Inputs coming from manufacturing industries use the price index in the database for that industry's shipments. Inputs coming from other (non-manufacturing) sectors use that sector's price index for gross output.

The main difficulty with implementing the above procedure is the lack of a clear one-to-one correspondence between the industry codes used in the use-make tables and those used in the ASM/CMF data. The use-make tables contain many fewer industries than the NAICS industries. BEA does provide concordance tables, which are supplemented with some manual matching based on industry definitions. We assume that all NAICS industries which map into the same use-make industry have the same structure of inputs.

To match MATCOST, this is a deflator for all materials, including energy. If you need a separate 'non-energy materials' deflator, you can construct an implicit one, based on (MATCOST/PIMAT) - (ENERGY/PIEN).