GTAP Data Bases: Domestic vs. Imported Goods
Below are some alternative suggestions. For best results, apply the procedure before changing from your original sectoral classification to the GTAP sectoral classification.- If your I-O table consists of a total use matrix and an imports matrix - Then just subtract the imports from total use to derive the domestic use matrix and then proceed in the usual way. This is the best scenario for the reporting of imports.
- If you only have a column vector reporting total imports by commodity - In this instance, we ask that you create an imports matrix by pro-rating the totals across uses by applying the structure implied by the total use matrix. For each row of the total use matrix, compute the percentage of the row total allocated to each sector. Then fill in the import matrix by multiplying each commodity total by the appropriate percentage for each sector. Finally, subtract the new imports matrix from the total use matrix to obtain the domestic use matrix. Under either this or the previous scenario, be sure to document the steps you followed to create the imports matrix.
- If you only have a row vector reporting total import costs by sector, but not by commodity - If this is how imports are treated in your I-O table, you need further information to prepare satisfactory import arrays. Specifically, you need data for imports by commodity. You also have a larger task than under the previous scenarios. The following paragraphs describe one way of performing this task.
Taking data on import usage by commodity, adjust it so that is consistent with your input-output data on import usage by use category. That is, make sure that the two import data sets give the same value for total imports. To achieve this, rescale the import usage by commodity data. Using these rescaled data, calculate import shares for domestic usage of each commodity (i.e. usage excluding exports). Apply these import shares to your input-output data for commodity usage by commodity and use category, to obtain initial estimates for import usage by commodity and use category. Using the RAS procedure, adjust these initial estimates to impose consistency with both the imports by commodity and the imports by use category data. To apply this procedure, first rescale each column of your initial estimates, to make it consistent with the imports by use category data. Then rescale each row, to make it consistent with the imports by commodity data. Continue rescaling columns and rows alternately, until you converge on a data set that satisfies both the imports by use category and the imports by commodity constraints simultaneously.
Last Modified: 3/5/2026 9:59:25 AM
